Dissemination

Dissemination

 

Y1 Y2 Y3 Y4 Total
Publications Journal articles 4 12 5 5 26
Book chapters 0 1 0 0 1
Conf. and workshop papers 5 25 27 14 71
PhD theses 1 1 0 1 3
Events Panels 1 0 0 0 1
Invited talks 0 2 5 5 12
Tutorials 0 0 0 1 1
Poster presentations 0 2 0 1 3
Patents 1 2 0 2 5
Software 2 1 0 1 4
Videos 0 2 0 2 4
Total 14 48 37 32 131

Publications

2017

  • Adrian Galdran, Javier Vazquez-Corral, David Pardo, and Marcelo Bertalmío. Fusion-Based Variational Image Dehazing. IEEE Signal Processing Letters, 24:151-155, 2017. doi:10.1109/LSP.2016.2643168
    [BibTeX]
    @Article{2017-02-GALDRAN,
    author = {Adrian Galdran and Javier Vazquez-Corral and David Pardo and Marcelo Bertalm\'{i}o},
    title = {{Fusion-Based Variational Image Dehazing}},
    journal = {{IEEE Signal Processing Letters}},
    date = {2017-02},
    doi = {10.1109/LSP.2016.2643168},
    volume = {24},
    issue = {2},
    year = {2017},
    pages = {151-155}
    }

2016

  • Angel Alvarez, Inigo Ugarte, Patricia Martinez, and Victor Fernandez. HW-SW Codesign of a Positioning System: From UML to Implementation Case Study. In Proceedings of the 2016 Conference on Design of Circuits and Integrated Systems (DCIS), pages 1-6, Grenada, Spain, 2016. doi:10.1109/DCIS.2016.7845378
    [BibTeX]
    @InProceedings{2016-11-ALVAREZ,
    author = {Angel Alvarez and Inigo Ugarte and Patricia Martinez and Victor Fernandez},
    title = {{HW-SW Codesign of a Positioning System: From UML to Implementation Case Study}},
    booktitle = {{Proceedings of the 2016 Conference on Design of Circuits and Integrated Systems (DCIS)}},
    date = {2016-11-23/2016-11-25},
    year = {2016},
    doi = {10.1109/DCIS.2016.7845378},
    address = {Grenada, Spain},
    pages = {1-6}
    }

  • Salma Bergaoui, Ivan Llopard, Nicolas Hili, Christian Fabre, and Fayçal Benaziz. Tool Support for a Method and a Language Integrating Model Refinements and Project Management. In Proceedings of the 8th Embedded Real-Time Software and Systems Congress (ERTS2 2016), pages 1-9, Toulouse, France, 2016.
    [BibTeX] [Download PDF]
    @Conference{2016-01-BERGAOUI,
    author = {Salma Bergaoui and Ivan Llopard and Nicolas Hili and Christian Fabre and Fay\c{c}al Benaziz},
    title = {{Tool Support for a Method and a Language Integrating Model Refinements and Project Management}},
    booktitle = {{Proceedings of the 8th Embedded Real-Time Software and Systems Congress (ERTS2 2016)}},
    address = {Toulouse, France},
    date = {2016-01-27/2016-01-29},
    year = {2016},
    pages = {1-9},
    url = {http://www.erts2016.org/uploads/pdf/erts_2016_Proceedings.pdf}
    }

  • Isah A. Lawal, Fabio Poiesi, Davide Anguita, and Andrea Cavallaro. Support Vector Motion Clustering. IEEE Transactions on Circuits and Systems for Video Technology, PP, 2016. doi:10.1109/TCSVT.2016.2580401
    [BibTeX] [Abstract]

    We present a closed-loop unsupervised clustering method for motion vectors extracted from highly dynamic video scenes. Motion vectors are assigned to non-convex homogeneous clusters characterizing direction, size and shape of regions with multiple independent activities. The proposed method is based on Support Vector Clustering (SVC). Cluster labels are propagated over time via incremental learning. The proposed method uses a kernel function that maps the input motion vectors into a high dimensional space to produce non-convex clusters. We improve the mapping effectiveness by quantifying feature similarities via a blend of position and orientation affinities. We use the Quasiconformal Kernel Transformation to boost the discrimination of outliers. The temporal propagation of the clusters? identities is achieved via incremental learning based on the concept of feature obsolescence to deal with appearing and disappearing features. Moreover, we design an on-line clustering performance prediction algorithm used as a feedback (closed-loop) that refines the cluster model at each frame in an unsupervised manner. We evaluate the proposed method on synthetic datasets and real-world crowded videos, and show that our solution outperforms state-of-the-art approaches.

    @article{2016-06-LAWAL,
    author = {Isah A. Lawal and Fabio Poiesi and Davide Anguita and Andrea Cavallaro},
    journal = {{IEEE Transactions on Circuits and Systems for Video Technology}},
    title = {{Support Vector Motion Clustering}},
    date = {2016-06-13},
    year = {2016},
    volume = {PP},
    issue = {99},
    doi = {10.1109/TCSVT.2016.2580401},
    abstract = {We present a closed-loop unsupervised clustering method for motion vectors extracted from highly dynamic video scenes. Motion vectors are assigned to non-convex homogeneous clusters characterizing direction, size and shape of regions with multiple independent activities. The proposed method is based on Support Vector Clustering (SVC). Cluster labels are propagated over time via incremental learning. The proposed method uses a kernel function that maps the input motion vectors into a high dimensional space to produce non-convex clusters. We improve the mapping effectiveness by quantifying feature similarities via a blend of position and orientation affinities. We use the Quasiconformal Kernel Transformation to boost the discrimination of outliers. The temporal propagation of the clusters? identities is achieved via incremental learning based on the concept of feature obsolescence to deal with appearing and disappearing features. Moreover, we design an on-line clustering performance prediction algorithm used as a feedback (closed-loop) that refines the cluster model at each frame in an unsupervised manner. We evaluate the proposed method on synthetic datasets and real-world crowded videos, and show that our solution outperforms state-of-the-art approaches.}
    }

  • Ivan Llopard. Programming Embedded Manycore: Refinement and Optimizing Compilation of a Parallel Action Language for Hierarchical State Machines. PhD thesis, 2016.
    [BibTeX] [Abstract]

    Modeling languages propose convenient abstractions and transformations to handle the complexity of today’s embedded systems. Based on the formalism of Hierarchical State Machine, they enable the expression of hierarchical control parallelism. However, they face two importants challenges when it comes to model data-intensive applications: no unified approach that also accounts for data-parallel actions; and no effective code optimization and generation flows. In this thesis, we propose a modeling language extended with parallel action semantics and hierarchical indexed-state machines suitable for computationally intensive applications. Together with its formal semantics, we present an optimizing model compiler aiming for the generation of efficient data-parallel implementations.

    @PhdThesis{2016-04-LLOPARD,
    author = {Ivan Llopard},
    title = {{Programming Embedded Manycore: Refinement and Optimizing Compilation of a Parallel Action Language for Hierarchical State Machines}},
    institution = {Universit\'e Pierre \& Marie Curie, Paris},
    date = {2016-04-26},
    year = {2016},
    abstract = {Modeling languages propose convenient abstractions and transformations to handle the complexity of today's embedded systems. Based on the formalism of Hierarchical State Machine, they enable the expression of hierarchical control parallelism. However, they face two importants challenges when it comes to model data-intensive applications: no unified approach that also accounts for data-parallel actions; and no effective code optimization and generation flows. In this thesis, we propose a modeling language extended with parallel action semantics and hierarchical indexed-state machines suitable for computationally intensive applications. Together with its formal semantics, we present an optimizing model compiler aiming for the generation of efficient data-parallel implementations.}
    }

  • Ivan Llopard, Christian Fabre, and Albert Cohen. From a Formalized Parallel Action Language to its Efficient Code Generation. ACM Transactions on Embedded Computing Systems, 16(2), 2016. doi:10.1145/2990195
    [BibTeX] [Abstract]

    Modeling languages propose convenient abstractions and transformations to handle the complexity of s embedded systems. Based on the formalism of Hierarchical State Machine, they enable the expression of hierarchical control parallelism. However, they face two importants challenges when it comes to model data-intensive applications: no unified approach that also accounts for data-parallel actions; and no effective code optimization and generation flows. We propose a modeling language extended with parallel action semantics and hierarchical indexed-state machines suitable for computationally intensive applications. Together with its formal semantics, we present an optimizing model compiler aiming for the generation of efficient data-parallel implementations.

    @Article{2016-09-LLOPARD,
    author = {Ivan Llopard and Christian Fabre and Albert Cohen},
    journal = {{ACM Transactions on Embedded Computing Systems}},
    title = {{From a Formalized Parallel Action Language to its Efficient Code Generation}},
    year = {2016},
    date = {2016-09},
    volume = {16},
    number = {2},
    doi = {10.1145/2990195},
    abstract = {Modeling languages propose convenient abstractions and transformations to handle the complexity of s embedded systems. Based on the formalism of Hierarchical State Machine, they enable the expression of hierarchical control parallelism. However, they face two importants challenges when it comes to model data-intensive applications: no unified approach that also accounts for data-parallel actions; and no effective code optimization and generation flows. We propose a modeling language extended with parallel action semantics and hierarchical indexed-state machines suitable for computationally intensive applications. Together with its formal semantics, we present an optimizing model compiler aiming for the generation of efficient data-parallel implementations.}
    }

  • Patricia Martinez and Eugenio Villar. Positioning System for Recreated Reality Applications Implemented on a Multi-Processing Embedded System. In I Jornadas de Computacion Empotrada y Reconfigurable (JCER 2016), Salamanca, Spain, 2016.
    [BibTeX]
    @InProceedings{2016-09-MARTINEZ,
    author = {Patricia Martinez and Eugenio Villar},
    title = {{Positioning System for Recreated Reality Applications Implemented on a Multi-Processing Embedded System}},
    booktitle = {{I Jornadas de Computacion Empotrada y Reconfigurable (JCER 2016)}},
    date = {2016-09-14/2016-09-16},
    year = {2016},
    address = {Salamanca, Spain}
    }

  • Ricardo Sanchez-Matilla, Fabio Poiesi, and Andrea Cavallaro. Online multi-target tracking with strong and weak detections. In Proceedings of European Conference on Computer Vision Workshop: 2nd Workshop on Benchmarking Multi-target Tracking: MOTChallenge 2016, pages 1-16, Amsterdam, The Netherlands, 2016.
    [BibTeX] [Abstract] [Download PDF]

    We propose an online multi-target tracker that exploits both high- and low-confidence target detections in a Probability Hypothesis Density Particle Filter framework. High-confidence (strong) detections are used for label propagation and target initialization. Low-confidence (weak) detections only support the propagation of labels, i.e. tracking existing targets. Moreover, we perform data association just after the prediction stage thus avoiding the need for computationally expensive labeling procedures such as clustering. Finally, we perform sampling by considering the perspective distortion in the target observations. The tracker runs on average at 12 frames per second. Results show that our method outperforms alternative online trackers on the Multiple Object Tracking 2016 and 2015 benchmark datasets in terms tracking accuracy, false negatives and speed.

    @InProceedings{2016-10-MATILLA,
    title = {{Online multi-target tracking with strong and weak detections}},
    author = {Ricardo Sanchez-Matilla and Fabio Poiesi and Andrea Cavallaro},
    booktitle = {{Proceedings of European Conference on Computer Vision Workshop: 2nd Workshop on Benchmarking Multi-target Tracking: MOTChallenge 2016}},
    address= {Amsterdam, The Netherlands},
    date = {2016-10-09},
    year = {2016},
    pages = {1-16},
    url = {http://www.eecs.qmul.ac.uk/~andrea/papers/2016_ECCVW_MOT_OnlineMTTwithStrongAndWeakDetections_Sanchez-Matilla_Poiesi_Cavallaro.pdf},
    abstract = {We propose an online multi-target tracker that exploits both high- and low-confidence target detections in a Probability Hypothesis Density Particle Filter framework. High-confidence (strong) detections are used for label propagation and target initialization. Low-confidence (weak) detections only support the propagation of labels, i.e. tracking existing targets. Moreover, we perform data association just after the prediction stage thus avoiding the need for computationally expensive labeling procedures such as clustering. Finally, we perform sampling by considering the perspective distortion in the target observations. The tracker runs on average at 12 frames per second. Results show that our method outperforms alternative online trackers on the Multiple Object Tracking 2016 and 2015 benchmark datasets in terms tracking accuracy, false negatives and speed.}
    }

  • Anca Molnos, Suzanne Lesecq, Julien Mottin, and Diego Puschini. Investigation of Q-learning applied to DVFS management of a System-on-Chip. In Proceedings of 4th IFAC International Conference on Intelligent Control and Automation Sciences, pages 278-284, Reims, France, 2016. doi:10.1016/j.ifacol.2016.07.126
    [BibTeX] [Abstract]

    This paper presents a new Q-learning based strategy to manage Dynamic Voltage Frequency Scaling (DVFS) on a system on chip (SoC) such that the energy consumption is reduced. We address software applications with throughput constraints. The proposed Q-learning formulation has two main advantages: it has a reduced state space to limit the overhead and it embeds a new reward function tailored for throughput-constrained applications. The DVFS manager is evaluated on a test chip executing an HMAX object recognition application. We perform an experimental investigation of the main Q-learning parameters. The results suggest that the proposed method reduces the energy consumed with up to 44% at the cost of occasionally increasing the number of throughput violations, when compared to two state-of-the-art feedback controllers that address the same application domain.

    @InProceedings{2016-06-MOLNOS,
    author = {Anca Molnos and Suzanne Lesecq and Julien Mottin and Diego Puschini},
    title = {{Investigation of Q-learning applied to DVFS management of a System-on-Chip}},
    booktitle = {{Proceedings of 4th IFAC International Conference on Intelligent Control and Automation Sciences}},
    date = {2016-06-01/2016-06-03},
    address = {Reims, France},
    pages = {278-284},
    year = {2016},
    doi = {10.1016/j.ifacol.2016.07.126},
    abstract = {This paper presents a new Q-learning based strategy to manage Dynamic Voltage Frequency Scaling (DVFS) on a system on chip (SoC) such that the energy consumption is reduced. We address software applications with throughput constraints. The proposed Q-learning formulation has two main advantages: it has a reduced state space to limit the overhead and it embeds a new reward function tailored for throughput-constrained applications. The DVFS manager is evaluated on a test chip executing an HMAX object recognition application. We perform an experimental investigation of the main Q-learning parameters. The results suggest that the proposed method reduces the energy consumed with up to 44% at the cost of occasionally increasing the number of throughput violations, when compared to two state-of-the-art feedback controllers that address the same application domain.}
    }

  • Raheeb Muzaffar, Vladimir Vukadinovic, and Andrea Cavallaro. Rate-adaptive Multicast Video Streaming from Teams of Micro Aerial Vehicles. In Proceedings of IEEE International Conference on Robotics and Automation, pages 1194-1201, Stockholm, Sweden, 2016. doi:10.1109/ICRA.2016.7487250
    [BibTeX] [Abstract]

    Video multicasting from cameras mounted on Micro Aerial Vehicles (MAVs) is desirable for applications such as search and rescue, surveillance and disaster management. Because of the mobility of the video sources and the high datarate of videos, the transmission rate should be adapted to the task at hand. Rate-adaptive video multicast streaming in 802.11 requires wireless link estimation as well as frequent feedback from multiple receivers. We propose an application layer rate-adaptive video multicast streaming framework using 802.11 adhoc network that is applicable when both the sender and the receiver nodes are mobile. The receiver nodes of a multicast group are dynamically elected based on their changing link conditions to gain feedback. An Application Layer Video Multicast Gateway (ALVM-GW) adapts the transmission rate and the video encoding rate based on the received feedback. Emulation results show that the proposed approach has balanced performance in terms of goodput, delay and packet loss.

    @InProceedings{2016-05-MUZAFFAR,
    title = {{Rate-adaptive Multicast Video Streaming from Teams of Micro Aerial Vehicles}},
    author = {Raheeb Muzaffar and Vladimir Vukadinovic and Andrea Cavallaro},
    booktitle = {{Proceedings of IEEE International Conference on Robotics and Automation}},
    address = {Stockholm, Sweden},
    date = {2016-05-16/2016-05-21},
    year = {2016},
    pages = {1194-1201},
    doi = {10.1109/ICRA.2016.7487250},
    abstract = {Video multicasting from cameras mounted on Micro Aerial Vehicles (MAVs) is desirable for applications such as search and rescue, surveillance and disaster management. Because of the mobility of the video sources and the high datarate of videos, the transmission rate should be adapted to the task at hand. Rate-adaptive video multicast streaming in 802.11 requires wireless link estimation as well as frequent feedback from multiple receivers. We propose an application layer rate-adaptive video multicast streaming framework using 802.11 adhoc network that is applicable when both the sender and the receiver nodes are mobile. The receiver nodes of a multicast group are dynamically elected based on their changing link conditions to gain feedback. An Application Layer Video Multicast Gateway (ALVM-GW) adapts the transmission rate and the video encoding rate based on the received feedback. Emulation results show that the proposed approach has balanced performance in terms of goodput, delay and packet loss.}
    }

  • Raheeb Muzaffar, Evsen Yanmaz, Christian Bettstetter, and Andrea Cavallaro. Application-layer rate-adaptive multicast video streaming over 802.11 for mobile devices. In Proceedings of the 24th ACM international conference on Multimedia, pages 506-510, Amsterdam, The Netherlands, 2016. doi:10.1145/2964284.2967272
    [BibTeX] [Abstract]

    Multicast video streaming over IEEE 802.11 is unreliable due to the lack of feedback from receivers. High data rates and variable link conditions require feedback from the receivers for link estimation to improve reliability and rate adaptation accordingly. In this paper, we validate on a test platform an application-layer rate-adaptive video multicast streaming framework using an 802.11 ad-hoc network applicable for mobile senders and receivers. Experimental results serve as a proof of concept and show the performance in terms of goodput, delay, packet loss, and received video quality.

    @InProceedings{2016-10-MUZAFFAR,
    title = {{Application-layer rate-adaptive multicast video streaming over 802.11 for mobile devices}},
    author = {Raheeb Muzaffar and Evsen Yanmaz and Christian Bettstetter and Andrea Cavallaro},
    booktitle = {{Proceedings of the 24th ACM international conference on Multimedia}},
    address= {Amsterdam, The Netherlands},
    date = {2016-10-15/2016-10-19},
    year = {2016},
    pages = {506-510},
    doi = {10.1145/2964284.2967272},
    abstract = {Multicast video streaming over IEEE 802.11 is unreliable due to the lack of feedback from receivers. High data rates and variable link conditions require feedback from the receivers for link estimation to improve reliability and rate adaptation accordingly. In this paper, we validate on a test platform an application-layer rate-adaptive video multicast streaming framework using an 802.11 ad-hoc network applicable for mobile senders and receivers. Experimental results serve as a proof of concept and show the performance in terms of goodput, delay, packet loss, and received video quality.}
    }

  • Jesper Puge Nielsen and Sven Karlsson. A scalable lock-free hash table with open addressing. In Proceedings of the 21st ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming (PPoPP 2016), pages 33:1-33:2, New York, NY, USA, 2016. ACM. doi:10.1145/2851141.2851196
    [BibTeX] [Abstract]

    Concurrent data structures synchronized with locks do not scale well with the number of threads. As more scalable alternatives, concurrent data structures and algorithms based on widely available, however advanced, atomic operations have been proposed. These data structures allow for correct and concurrent operations without any locks. In this paper, we present a new fully lock-free open addressed hash table with a simpler design than prior published work. We split hash table insertions into two atomic phases: first inserting a value ignoring other concurrent operations, then in the second phase resolve any duplicate or conflicting values. Our hash table has a constant and low memory usage that is less than existing lock-free hash tables at a fill level of 33\% and above. The hash table exhibits good cache locality. Compared to prior art, our hash table results in 16\% and 15\% fewer L1 and L2 cache misses respectively, leading to 21\% fewer memory stall cycles. Our experiments show that our hash table scales close to linearly with the number of threads and outperforms, in throughput, other lock-free hash tables by 19\%.

    @InProceedings{2016-03-NIELSEN,
    title = {{A scalable lock-free hash table with open addressing}},
    author = {Jesper Puge Nielsen and Sven Karlsson},
    booktitle = {{Proceedings of the 21st ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming (PPoPP 2016)}},
    date = {2016-03-12/2016-03-16},
    year = {2016},
    pages = {33:1-33:2},
    doi = {10.1145/2851141.2851196},
    date = {2016},
    publisher = {ACM},
    address = {New York, NY, USA},
    abstract = {Concurrent data structures synchronized with locks do not scale well with the number of threads. As more scalable alternatives, concurrent data structures and algorithms based on widely available, however advanced, atomic operations have been proposed. These data structures allow for correct and concurrent operations without any locks. In this paper, we present a new fully lock-free open addressed hash table with a simpler design than prior published work. We split hash table insertions into two atomic phases: first inserting a value ignoring other concurrent operations, then in the second phase resolve any duplicate or conflicting values.
    Our hash table has a constant and low memory usage that is less than existing lock-free hash tables at a fill level of 33\% and above. The hash table exhibits good cache locality. Compared to prior art, our hash table results in 16\% and 15\% fewer L1 and L2 cache misses respectively, leading to 21\% fewer memory stall cycles. Our experiments show that our hash table scales close to linearly with the number of threads and outperforms, in throughput, other lock-free hash tables by 19\%.}
    }

  • Ayoub Nouri, Rahma Ben Atitallah, Anca Molnos, Christian Fabre, Frédéric Heitzmann, and Olivier Debicki. Transforming VHDL Descriptions into Formal Component-based Models. In Proceedings of 27th IEEE International Symposium on Rapid System Prototyping (RSP), Pittsburgh, PA, USA, 2016. To appear doi:10.1145/2990299.2990320
    [BibTeX]
    @InProceedings{2016-10-NOURI,
    author = {Ayoub Nouri and Rahma Ben Atitallah and Anca Molnos and Christian Fabre and Fr\'{e}d\'{e}ric Heitzmann and Olivier Debicki},
    title = {{Transforming VHDL Descriptions into Formal Component-based Models}},
    booktitle = {{Proceedings of 27th IEEE International Symposium on Rapid System Prototyping (RSP)}},
    date = {2016-10-06/2016-10-07},
    year = {2016},
    doi = {10.1145/2990299.2990320},
    note = {To appear},
    address = {Pittsburgh, PA, USA}
    }

  • Fabio Poiesi and Andrea Cavallaro. Detection of fast incoming objects with a moving camera. In British Machine Vision Conference, pages 1-11, York, UK, 2016.
    [BibTeX] [Abstract] [Download PDF]

    Using a monocular camera for early collision detection in cluttered scenes to elude fast incoming objects is a desirable but challenging functionality for mobile robots, such as small drones. We present a novel moving object detection and avoidance algorithm for an uncalibrated camera that uses only the optical flow to predict collisions. First, we estimate the optical flow and compensate the global camera motion. Then we detect incoming objects while removing the noise caused by dynamic textures, nearby terrain and lens distortion by means of an adaptively learnt background-motion model. Next, we estimate the time to contact, namely the expected time for an incoming object to cross the infinite plane defined by the extension of the image plane. Finally, we combine the time to contact and the compensated motion in a Bayesian framework to identify an object-free region the robot can move towards to avoid the collision. We demonstrate and evaluate the proposed algorithm using footage of flying robots that observe fast incoming objects such as birds, balls and other drones.

    @InProceedings{2016-09-POIESI,
    title = {{Detection of fast incoming objects with a moving camera}},
    author = {Fabio Poiesi and Andrea Cavallaro},
    booktitle = {{British Machine Vision Conference}},
    address= {York, UK},
    date = {2016-09-19/2016-09-22},
    year = {2016},
    pages = {1-11},
    url = {http://fabio-poiesi.com/files/papers/conferences/2016_BMVC_DetectionFastIncomingObjects_Poiesi_Cavallaro.pdf},
    abstract = {Using a monocular camera for early collision detection in cluttered scenes to elude fast incoming objects is a desirable but challenging functionality for mobile robots, such as small drones. We present a novel moving object detection and avoidance algorithm for an uncalibrated camera that uses only the optical flow to predict collisions. First, we estimate the optical flow and compensate the global camera motion. Then we detect incoming objects while removing the noise caused by dynamic textures, nearby terrain and lens distortion by means of an adaptively learnt background-motion model. Next, we estimate the time to contact, namely the expected time for an incoming object to cross the infinite plane defined by the extension of the image plane. Finally, we combine the time to contact and the compensated motion in a Bayesian framework to identify an object-free region the robot can move towards to avoid the collision. We demonstrate and evaluate the proposed algorithm using footage of flying robots that observe fast incoming objects such as birds, balls and other drones.}
    }

  • Mateusz Rzymowski, Przemys textbarl, and L. Single-Anchor Indoor Localization Using ESPAR Antenna. IEEE Antennas and Wireless Propagation Letters, 15:1183-1186, 2016. doi:10.1109/LAWP.2015.2498950
    [BibTeX]

  • Mateusz Rzymowski, Damian Duraj, Ł., Krzysztof Nyka, and Przemysł. UHF ESPAR Antenna for Simple Angle of Arrival Estimation in UHF RFID Applications. In Proceedings of Microwaves, Radar, and Wireless Communication (MIKON) 21th IEEE International Conference, pages 1-4, Krakow, Poland, 2016. doi:10.1109/MIKON.2016.7491984
    [BibTeX] [Abstract]

    An electronically switched beam antenna for localization of passive UHF RFID tags based on a simple Angle of Arrival (AoA) technique is proposed. Detailed antenna design and realization are presented together with corresponding simulations and measurement results. Experimental tests with passive UHF RFID tag show the validity of theoretical assumptions for application of the antenna in UHF RFID based localization systems.

    @InProceedings{2016-05-RZYMOWSKI,
    author = {Mateusz Rzymowski and Damian Duraj and \L{}ukasz Kulas and Krzysztof Nyka and Przemys\l{}aw Wo\'{z}nica},
    title = {{UHF ESPAR Antenna for Simple Angle of Arrival Estimation in UHF RFID Applications}},
    booktitle = {{Proceedings of Microwaves, Radar, and Wireless Communication (MIKON) 21th IEEE International Conference}},
    date = {2016-05-09/2016-05-11},
    year = {2016},
    pages = {1-4},
    doi = {10.1109/MIKON.2016.7491984},
    address = {Krakow, Poland},
    abstract = {An electronically switched beam antenna for localization of passive UHF RFID tags based on a simple Angle of Arrival (AoA) technique is proposed. Detailed antenna design and realization are presented together with corresponding simulations and measurement results. Experimental tests with passive UHF RFID tag show the validity of theoretical assumptions for application of the antenna in UHF RFID based localization systems.}
    }

  • Juan C. SanMiguel and Andrea Cavallaro. Energy consumption models for smart-camera networks. IEEE Transactions on Circuits and Systems for Video Technology, 99:1-14, 2016. doi:10.1109/TCSVT.2016.2593598
    [BibTeX] [Abstract]

    Camera networks require heavy visual-data processing and high-bandwidth communication. In this paper, we identify key factors underpinning the development of resourceaware algorithms and we propose a comprehensive energy consumption model for the resources employed by smart-camera networks, which are composed of cameras that process data locally and collaborate with their neighbours. We account for the main parameters that influence consumption when sensing (framesize and framerate), processing (dynamic frequency scaling and task load) and communication (output power and bandwidth) are considered. Next we define an abstraction based on clock frequency and duty cycle that accounts for active, idle and sleep operational states. We demonstrate the importance of the proposed model for a multi-camera tracking task and show how one may significantly reduce consumption with only minor performance degradation when choosing to operate with an appropriately reduced hardware capacity. Moreover, we quantify the dependency on local computation resources and on bandwidth availability. The proposed consumption model can be easily adjusted to account for new platforms, thus providing a valuable tool for the design of resource-aware algorithms and further research in resource-aware camera networks.

    @Article{2016-07-SANMIGUEL,
    author = {Juan C. SanMiguel and Andrea Cavallaro},
    journal = {{IEEE Transactions on Circuits and Systems for Video Technology}},
    title = {{Energy consumption models for smart-camera networks}},
    date = {2016-07-20},
    year = {2016},
    volume = {99},
    pages = {1-14},
    doi = {10.1109/TCSVT.2016.2593598},
    abstract = {Camera networks require heavy visual-data processing and high-bandwidth communication. In this paper, we identify key factors underpinning the development of resourceaware algorithms and we propose a comprehensive energy consumption model for the resources employed by smart-camera networks, which are composed of cameras that process data locally and collaborate with their neighbours. We account for the main parameters that influence consumption when sensing (framesize and framerate), processing (dynamic frequency scaling and task load) and communication (output power and bandwidth) are considered. Next we define an abstraction based on clock frequency and duty cycle that accounts for active, idle and sleep operational states. We demonstrate the importance of the proposed model for a multi-camera tracking task and show how one may significantly reduce consumption with only minor performance degradation when choosing to operate with an appropriately reduced hardware capacity. Moreover, we quantify the dependency on local computation resources and on bandwidth availability. The proposed consumption model can be easily adjusted to account for new platforms, thus providing a valuable tool for the design of resource-aware algorithms and further research in resource-aware camera networks.}
    }

  • Piotr Szczuko. Simple Gait Parameterization and 3D Animation for Anonymous Visual Monitoring Based on Augmented Reality. Multimedia Tools and Applications, 75(17):10561-10581, 2016. doi:10.1007/s11042-015-2874-0
    [BibTeX] [Abstract]

    The article presents a method for video anonymization and replacing real human silhouettes with virtual 3D figures rendered on a screen. Video stream is processed to detect and to track objects, whereas anonymization stage employs animating avatars accordingly to behavior of detected persons. Location, movement speed, direction, and person height are taken into account during animation and rendering phases. This approach requires a calibrated camera, and utilizes results of visual object tracking. A procedure for transforming objects visual features and bounding boxes into gait parameters for animated figures is presented. Conclusions and future work perspectives are provided.

    @Article{2016-09-SZCZUKOa,
    author = {Piotr Szczuko},
    title = {{Simple Gait Parameterization and 3D Animation for Anonymous Visual Monitoring Based on Augmented Reality}},
    journal = {{Multimedia Tools and Applications}},
    volume = {75},
    number = {17},
    date = {2016-09-01},
    pages = {10561--10581},
    doi = {10.1007/s11042-015-2874-0},
    issn = {1380-7501},
    publisher = {Springer},
    abstract = {The article presents a method for video anonymization and replacing real human silhouettes with virtual 3D figures rendered on a screen. Video stream is processed to detect and to track objects, whereas anonymization stage employs animating avatars accordingly to behavior of detected persons. Location, movement speed, direction, and person height are taken into account during animation and rendering phases. This approach requires a calibrated camera, and utilizes results of visual object tracking. A procedure for transforming objects visual features and bounding boxes into gait parameters for animated figures is presented. Conclusions and future work perspectives are provided.},
    year = {2016}
    }

  • Piotr Szczuko, Grzegorz Szwoch, Maciej Szczodrak, Jozef Kotus, and Andrzej Czyzewski. Zastosowania dronow i sensorow wizyjnych i akustycznych do zdalnej detekcji i lokalizacji obiektow i zdarzen (Application of Drones and Acoustic and Visual Sensors in Remote Detection and Localization of Objects and Events). In Krajowe Sympozjum Telekomunikacji i Teleinformatyki (KSTiT 2016), Gliwice, Poland, 2016.
    [BibTeX]
    @InProceedings{2016-09-SZCZUKOb,
    author = {Piotr Szczuko and Grzegorz Szwoch and Maciej Szczodrak and Jozef Kotus and Andrzej Czyzewski},
    title = {{Zastosowania dronow i sensorow wizyjnych i akustycznych do zdalnej detekcji i lokalizacji obiektow i zdarzen (Application of Drones and Acoustic and Visual Sensors in Remote Detection and Localization of Objects and Events)}},
    booktitle = {{Krajowe Sympozjum Telekomunikacji i Teleinformatyki (KSTiT 2016)}},
    address = {Gliwice, Poland},
    date = {2016-09-26/2016-09-28},
    year = {2016}
    }

  • Grzegorz Szwoch. Performance evaluation of the parallel object tracking algorithm employing the particle filter. In 20th IEEE Conference SPA 2016, Signal Processing, Algorithms, Architectures, Arrangements, and Applications, Poznan, Poland, 2016. doi:10.1109/SPA.2016.7763598
    [BibTeX]
    @InProceedings{2016-09-SZWOCH,
    author = {Grzegorz Szwoch},
    title = {{Performance evaluation of the parallel object tracking algorithm employing the particle filter}},
    booktitle = {{20th IEEE Conference SPA 2016, Signal Processing, Algorithms, Architectures, Arrangements, and Applications}},
    date = {2016-09-21/2016-09-23},
    year = {2016},
    doi = {10.1109/SPA.2016.7763598},
    address = {Poznan, Poland}
    }

  • Tea Tušar, Klemen Gantar, and Bogdan Filipič. The pitfalls of overfitting in optimization of a manufacturing quality control procedure. In Proceedings of the 7th International Conference on Bioinspired Optimization Methods and their Applications, BIOMA 2016, pages 241-253, Bled, Slovenia, 2016.
    [BibTeX] [Abstract] [Download PDF]

    We are concerned with the estimation of copper-graphite joints quality in commutator manufacturing—a classification problem in which we wish to detect whether the joints are soldered well or have any of the four known defects. This quality control procedure can be automated by means of an on-line classifier that can assess the quality of commutators as they are being manufactured. A classifier suitable for this task can be constructed by combining computer vision, machine learning and evolutionary optimization techniques. While previous work has shown the validity of this approach, this paper demonstrates that the search for an accurate classifier can lead to overfitting despite cross-validation being used for assessing the classifier performance. We inspect several aspects of this phenomenon and propose to use repeated cross-validation in order to amend it.

    @InProceedings{2016-05-TUSAR,
    title = {{The pitfalls of overfitting in optimization of a manufacturing quality control procedure}},
    author = {Tea Tu\v{s}ar and Klemen Gantar and Bogdan Filipi\v{c}},
    booktitle = {{Proceedings of the 7th International Conference on Bioinspired Optimization Methods and their Applications, BIOMA 2016}},
    pages = {241-253},
    date = {2016-05-18/2016-05-20},
    year = {2016},
    address = {Bled, Slovenia},
    url = {http://bioma.ijs.si/proceedings/2016/17%20-%20The%20Pitfalls%20of%20Overfitting%20in%20Optimization%20of%20a%20Manufacturing%20Quality%20Control%20Procedure.pdf},
    abstract = {We are concerned with the estimation of copper-graphite joints quality in commutator manufacturing---a classification problem in which we wish to detect whether the joints are soldered well or have any of the four known defects. This quality control procedure can be automated by means of an on-line classifier that can assess the quality of commutators as they are being manufactured. A classifier suitable for this task can be constructed by combining computer vision, machine learning and evolutionary optimization techniques. While previous work has shown the validity of this approach, this paper demonstrates that the search for an accurate classifier can lead to overfitting despite cross-validation being used for assessing the classifier performance. We inspect several aspects of this phenomenon and propose to use repeated cross-validation in order to amend it.}
    }

  • Sven Verdoolaege and Albert Cohen. Live Range Reordering. In 6th Workshop on Polyhedral Compilation Techniques (IMPACT), Prague, Czech Republic, 2016. Associated with HiPEAC 2016
    [BibTeX] [Abstract] [Download PDF]

    False dependences are caused by the reuse of memory to store different data. These false dependences severely constrain the schedule of statement instances, effectively serializing successive accesses to the same memory location. Several array expansion techniques have been proposed to eliminate some or all of these false dependences, enabling more reorderings of statement instances during loop nest transformations. However, array expansion is only relevant when complemented with a storage mapping optimization step, typically taking advantage of the fixed schedule set in earlier phases of the compilation, folding successive values into a compact set of contracted arrays. Furthermore, array expansion can result in memory footprint and locality damages that may not be recoverable through storage mapping optimization when intermediate transformation steps have abused the freedom offered by the removal of false dependences. Array expansion and storage mapping optimization are also complex procedures not found in most compilers, and the latter is moreover performed using suboptimal heuristics (particularly in the multi-array case). Finally, array expansion may not remove all false dependences when considering data-dependent control and access patterns. For all these reasons, it is desirable to explore alternatives to array expansion as a means to avoid the spurious serialization effect of false dependences. This serialization is unnecessary in general, as semantics preservation in presence of memory reuse only requires the absence of interference among live-ranges, an unordered constraint compatible with the their commutation. We present a technique to deal with memory reuse without serializing successive uses of memory, but also without increasing memory requirements or preventing important loop transformations such as loop distribution. The technique is generic, fine-grained (instancewise) and extends two recently proposed, more restrictive approaches. It has been systematically tested in PPCG and shown to be essential to the parallelizing compilation of a variety of loop nests, including large pencil programs with many scalar variables.

    @InProceedings{2016-01-VERDOOLAEGE,
    author = {Sven Verdoolaege and Albert Cohen},
    title = {{Live Range Reordering}},
    booktitle = {{6th Workshop on Polyhedral Compilation Techniques (IMPACT)}},
    note = {Associated with HiPEAC 2016},
    year = {2016},
    date = {2016-01-19},
    address = {Prague, Czech Republic},
    url = {https://lirias.kuleuven.be/bitstream/123456789/519489/1/live_range_reordering.pdf},
    abstract = {False dependences are caused by the reuse of memory to store different data. These false dependences severely constrain the schedule of statement instances, effectively serializing successive accesses to the same memory location. Several array expansion techniques have been proposed to eliminate some or all of these false dependences, enabling more reorderings of statement instances during loop nest transformations. However, array expansion is only relevant when complemented with a storage mapping optimization step, typically taking advantage of the fixed schedule set in earlier phases of the compilation, folding successive values into a compact set of contracted arrays. Furthermore, array expansion can result in memory footprint and locality damages that may not be recoverable through storage mapping optimization when intermediate transformation steps have abused the freedom offered by the removal of false dependences. Array expansion and storage mapping optimization are also complex procedures not found in most compilers, and the latter is moreover performed using suboptimal heuristics (particularly in the multi-array case). Finally, array expansion may not remove all false dependences when considering data-dependent control and access patterns. For all these reasons, it is desirable to explore alternatives to array expansion as a means to avoid the spurious serialization effect of false dependences. This serialization is unnecessary in general, as semantics preservation in presence of memory reuse only requires the absence of interference among live-ranges, an unordered constraint compatible with the their commutation. We present a technique to deal with memory reuse without serializing successive uses of memory, but also without increasing memory requirements or preventing important loop transformations such as loop distribution. The technique is generic, fine-grained (instancewise) and extends two recently proposed, more restrictive approaches. It has been systematically tested in PPCG and shown to be essential to the parallelizing compilation of a variety of loop nests, including large pencil programs with many scalar variables.}
    }

  • Lin Wang and Andrea Cavallaro. Ear in the sky: Ego-noise reduction for auditory micro aerial vehicles. In Proceedings of 13th IEEE International Conference on Advanced Signal and Video based Surveillance (AVSS), Colorado Springs, USA, 2016. doi:10.1109/AVSS.2016.7738063
    [BibTeX] [Abstract]

    We investigate the spectral and spatial characteristics of the ego-noise of a multirotor micro aerial vehicle (MAV) using audio signals captured with multiple onboard microphones and derive a noise model that grounds the feasibility of microphone-array techniques for noise reduction. The spectral analysis suggests that the ego-noise consists of narrowband harmonic noise and broadband noise, whose spectra vary dynamically with the motor rotation speed. The spatial analysis suggests that the ego-noise of a P-rotor MAV can be modeled as P directional noises plus one diffuse noise. Moreover, because of the fixed positions of the microphones and motors, we can assume that the acoustic mixing network of the ego-noise is stationary. We validate the proposed noise model and the stationary mixing assumption by applying blind source separation to multi-channel recordings from both a static and a moving MAV and quantify the signal-to-noise ratio improvement. Moreover, we make all the audio recordings publicly available.

    @InProceedings{2016-08-WANG,
    title = {{Ear in the sky: Ego-noise reduction for auditory micro aerial vehicles}},
    author = {Lin Wang and Andrea Cavallaro},
    booktitle = {{Proceedings of 13th IEEE International Conference on Advanced Signal and Video based Surveillance (AVSS)}},
    address= {Colorado Springs, USA},
    date = {2016-08-23/2016-08-26},
    doi = {10.1109/AVSS.2016.7738063},
    year = {2016},
    abstract = {We investigate the spectral and spatial characteristics of the ego-noise of a multirotor micro aerial vehicle (MAV) using audio signals captured with multiple onboard microphones and derive a noise model that grounds the feasibility of microphone-array techniques for noise reduction. The spectral analysis suggests that the ego-noise consists of narrowband harmonic noise and broadband noise, whose spectra vary dynamically with the motor rotation speed. The spatial analysis suggests that the ego-noise of a P-rotor MAV can be modeled as P directional noises plus one diffuse noise. Moreover, because of the fixed positions of the microphones and motors, we can assume that the acoustic mixing network of the ego-noise is stationary. We validate the proposed noise model and the stationary mixing assumption by applying blind source separation to multi-channel recordings from both a static and a moving MAV and quantify the signal-to-noise ratio improvement. Moreover, we make all the audio recordings publicly available.}
    }

2015

  • Metin Aktas, Toygar Akgün, Duygu Büyükaydin, and Hüseyin Özkan. Acoustic Direciton Finding Under High Reverberation (Yüksek Yanilama Altinda Akustik Yön Bulma). In Proceedings of the 23nd IEEE Signal Processing and Communications Applications Conference (SIU 2015), pages 1533-1536, Inönü University, Malatya, Turkey, 2015. doi:10.1109/SIU.2015.7130138
    [BibTeX] [Abstract]

    One of the major challanges for acoustic source localization is to eliminate the multipath distortions due to the reverberation. In this paper, we propose a novel direction finding method that is robust to multipath distortions. Since we do not make any stochastic and sparseness assumptions regarding the underlying signal source, our method can be applied to a wide range of wideband acoustic signals. In particular, we introduce a novel approach to extract the signal portions that are not distorted with multipath signals and noise. Hence, the DOA estimation can be performed without being affected from the reverberation. Simulation results show that the proposed method reliably estimates the source direction even under the significantly high reverberation conditions.

    @InProceedings{2015-05-AKTAS,
    author = {Metin Aktas and Toygar Akg\"{u}n and Duygu B\"{u}y\"{u}kaydin and H\"{u}seyin \"{O}zkan},
    title = {{Acoustic Direciton Finding Under High Reverberation (Y\"{u}ksek Yanilama Altinda Akustik Y\"{o}n Bulma)}},
    booktitle = {{Proceedings of the 23nd IEEE Signal Processing and Communications Applications Conference (SIU 2015)}},
    date = {2015-05-16/2015-05-19},
    address = {In\"{o}n\"{u} University, Malatya, Turkey},
    doi = {10.1109/SIU.2015.7130138},
    pages = {1533-1536},
    abstract = {One of the major challanges for acoustic source localization is to eliminate the multipath distortions due to the reverberation. In this paper, we propose a novel direction finding method that is robust to multipath distortions. Since we do not make any stochastic and sparseness assumptions regarding the underlying signal source, our method can be applied to a wide range of wideband acoustic signals. In particular, we introduce a novel approach to extract the signal portions that are not distorted with multipath signals and noise. Hence, the DOA estimation can be performed without being affected from the reverberation. Simulation results show that the proposed method reliably estimates the source direction even under the significantly high reverberation conditions.},
    year = {2015}
    }

  • Metin Aktas, Toygar Akgün, and Hüseyin Özkan. Acoustic Direction Finding In Highly Reverberant Environment With Single Acoustic Vector Sensor. In Proceedings of the 23rd European Signal Processing Conference (EUSIPCO 2015), pages 2301-2305, Nice, France, 2015. doi:10.1109/EUSIPCO.2015.7362795
    [BibTeX] [Abstract]

    We propose a novel wideband acoustic direction finding method for highly reverberant environments using measurements from a single Acoustic Vector Sensor (AVS). Since an AVS is small in size and can be effectively used within the full acoustic frequency bands, the proposed solution is suitable for wideband acoustic source localization. In particular, we introduce a novel approach to extract the signal portions that are not distorted with multipath signals and noise. We do not make any stochastic and sparseness assumptions regarding the underlying signal source. Hence, our approach can be applied to a wide range of wideband acoustic signals. We present experiments with acoustic signals that are specially exposed to long reverberations, where the Signal-to-Noise Ratio is as low as 0 dB. In these experiments, the proposed method reliably estimates the source direction with less than 5 degrees of error even under the introduced significantly high reverberation conditions.

    @InProceedings{2015-08-AKTAS,
    author = {Metin Aktas and Toygar Akg\"{u}n and H\"{u}seyin \"{O}zkan},
    title = {{Acoustic Direction Finding In Highly Reverberant Environment With Single Acoustic Vector Sensor}},
    booktitle = {{Proceedings of the 23rd European Signal Processing Conference (EUSIPCO 2015)}},
    date = {2015-08-31/2015-09-04},
    pages = {2301-2305},
    doi = {10.1109/EUSIPCO.2015.7362795},
    address = {Nice, France},
    abstract = {We propose a novel wideband acoustic direction finding method for highly reverberant environments using measurements from a single Acoustic Vector Sensor (AVS). Since an AVS is small in size and can be effectively used within the full acoustic frequency bands, the proposed solution is suitable for wideband acoustic source localization. In particular, we introduce a novel approach to extract the signal portions that are not distorted with multipath signals and noise. We do not make any stochastic and sparseness assumptions regarding the underlying signal source. Hence, our approach can be applied to a wide range of wideband acoustic signals. We present experiments with acoustic signals that are specially exposed to long reverberations, where the Signal-to-Noise Ratio is as low as 0 dB. In these experiments, the proposed method reliably estimates the source direction with less than 5 degrees of error even under the introduced significantly high reverberation conditions.},
    year = {2015}
    }

  • Angelos Antonopoulos, Aris S. Lalos, Marco Di Renzo, and Christos Verikoukis. Cross-layer Theoretical Analysis of NC-aided Cooperative ARQ Protocols in Correlated Shadowed Environments. IEEE Transactions on Vehicular Technology, 64(9):4074-4087, 2015. doi:10.1109/TVT.2014.2361670
    [BibTeX]
    @Article{2015-09-ANTONOPOULOS,
    author = {Angelos Antonopoulos and Aris S. Lalos and Marco Di Renzo and Christos Verikoukis},
    title = {{Cross-layer Theoretical Analysis of NC-aided Cooperative ARQ Protocols in Correlated Shadowed Environments}},
    journal = {{IEEE Transactions on Vehicular Technology}},
    date = {2015-09},
    volume = {64},
    number = {9},
    pages = {4074-4087},
    year = {2015},
    doi = {10.1109/TVT.2014.2361670}
    }

  • Riyadh Baghdadi, Ulysse Beaugnon, Albert Cohen, Tobias Grosser, Michael Kruse, Chandan Reddy, Sven Verdoolaege, Mohammed Javed Absar, Sven Van Haastregt, Alexey Kravets, Anton Lokhmotov, Robert David, Elnar Hajiyev, Adam Betts, Alastair Donaldson, and Jeroen Ketema. PENCIL: a Platform-Neutral Compute Intermediate Language for Accelerator Programming. In Proceedings of The 24th International Conference on Parallel Architectures and Compilation Techniques (PACT 2015), pages 138-149, San Francisco, California, USA, 2015. doi:10.1109/PACT.2015.17
    [BibTeX] [Download PDF]

  • Lars Bonnichsen and Artur Podobas. Using Transactional Memory to Avoid Blocking in OpenMP Synchronization Directives. In Christian Terboven, Bronis R. de Supinski, Pablo Reble, Barbara M. Chapman, and Matthias S. Müller, editors, OpenMP: Heterogenous Execution and Data Movements, Proceedings of the 11th International Workshop on OpenMP (IWOMP), pages 149-161, Aachen, Germany, 2015. Springer. doi:10.1007/978-3-319-24595-9_11
    [BibTeX] [Abstract]

    OpenMP applications with abundant parallelism are often characterized by their high-performance. Unfortunately, OpenMP applications with a lot of synchronization or serialization-points perform poorly because of blocking, i.e. the threads have to wait for each other. In this paper, we present methods based on hardware transactional memory (HTM) for executing OpenMP barrier, critical, and taskwait directives without blocking. Although HTM is still relatively new in the Intel and IBM architectures, we experimentally show a 73 % performance improvement over traditional locking approaches, and 23 % better than other HTM approaches on critical sections. Speculation over barriers can decrease execution time by up-to 41 %. We expect that future systems with HTM support and more cores will have a greater benefit from our approach as they are more likely to block.

    @InProceedings{2015-10-BONNICHSEN,
    author = {Lars Bonnichsen and Artur Podobas},
    editor = {Christian Terboven and Bronis R. de Supinski and Pablo Reble and Barbara M. Chapman and Matthias S. M{\"u}ller},
    title = {{Using Transactional Memory to Avoid Blocking in OpenMP Synchronization Directives}},
    booktitle = {{OpenMP: Heterogenous Execution and Data Movements, Proceedings of the 11th International Workshop on OpenMP (IWOMP)}},
    date = {2015-10-01/2015-10-02},
    year = {2015},
    publisher = {Springer},
    address = {Aachen, Germany},
    pages = {149-161},
    doi = {10.1007/978-3-319-24595-9_11},
    abstract = {OpenMP applications with abundant parallelism are often characterized by their high-performance. Unfortunately, OpenMP applications with a lot of synchronization or serialization-points perform poorly because of blocking, i.e. the threads have to wait for each other. In this paper, we present methods based on hardware transactional memory (HTM) for executing OpenMP barrier, critical, and taskwait directives without blocking. Although HTM is still relatively new in the Intel and IBM architectures, we experimentally show a 73 % performance improvement over traditional locking approaches, and 23 % better than other HTM approaches on critical sections. Speculation over barriers can decrease execution time by up-to 41 %. We expect that future systems with HTM support and more cores will have a greater benefit from our approach as they are more likely to block.}
    }

  • Marie-Anne Burcklen, Frederic Diaz, François Leprêtre, Joel Rollin, Anne Delboulbé, Mane-Si Laure Lee, Brigitte Loiseaux, Allan Koudoli, Simon Denel, Philippe Millet, Francois Duhem, Fabrice Lemonnier, Hervé Sauer, and Francois Goudail. Experimental demonstration of extended depth-of-field f/1.2 visible High Definition camera with jointly optimized phase mask and real-time digital processing. Journal of the european optical society – rapid publications, 10:1-6, 2015. doi:10.2971/jeos.2015.15046
    [BibTeX] [Abstract]

    Increasing the depth of field (DOF) of compact visible high resolution cameras while maintaining high imaging performance in the DOF range is crucial for such applications as night vision goggles or industrial inspection. In this paper, we present thppe end-to-end design and experimental validation of an extended depth-of-field visible High Definition camera with a very small f-number, combining a six-ring pyramidal phase mask in the aperture stop of the lens with a digital deconvolution. The phase mask and the deconvolution algorithm are jointly optimized during the design step so as to maximize the quality of the deconvolved image over the DOF range. The deconvolution processing is implemented in real-time on a Field-Programmable Gate Array and we show that it requires very low power consumption. By mean of MTF measurements and imaging experiments we experimentally characterize the performance of both cameras with and without phase mask and thereby demonstrate a significant increase in depth of field of a factor 2.5, as it was expected in the design step.

    @Article{2015-10-BURCKLEN,
    author = {Marie-Anne Burcklen and Frederic Diaz and Fran\c{c}ois Lepr\^{e}tre and Joel Rollin and Anne Delboulb\'e and Mane-Si Laure Lee and Brigitte Loiseaux and Allan Koudoli and Simon Denel and Philippe Millet and Francois Duhem and Fabrice Lemonnier and Herv\'e Sauer and Francois Goudail},
    title = {{Experimental demonstration of extended depth-of-field f/1.2 visible High Definition camera with jointly optimized phase mask and real-time digital processing}},
    journal = {Journal of the European Optical Society - Rapid publications},
    date = {2015-10-21},
    year = {2015},
    volume = {10},
    pages = {1-6},
    doi = {10.2971/jeos.2015.15046},
    publisher = {EOS},
    abstract = {Increasing the depth of field (DOF) of compact visible high resolution cameras while maintaining high imaging performance in the DOF range is crucial for such applications as night vision goggles or industrial inspection. In this paper, we present thppe end-to-end design and experimental validation of an extended depth-of-field visible High Definition camera with a very small f-number, combining a six-ring pyramidal phase mask in the aperture stop of the lens with a digital deconvolution. The phase mask and the deconvolution algorithm are jointly optimized during the design step so as to maximize the quality of the deconvolved image over the DOF range. The deconvolution processing is implemented in real-time on a Field-Programmable Gate Array and we show that it requires very low power consumption. By mean of MTF measurements and imaging experiments we experimentally characterize the performance of both cameras with and without phase mask and thereby demonstrate a significant increase in depth of field of a factor 2.5, as it was expected in the design step.}
    }

  • Janusz Cichowski, Lisowski Karol, Szczuko Piotr, and Andrzej Czyżewski. Zdalny zintegrowany moduł nadzoru radiowo-wizyjnego (An integrated module for remote radio and visual monitoring). In Krajowe Sympozjum Telekomunikacji i Teleinformatyki (KSTiT 2015), Kraków, Poland, 2015. doi:10.15199/59.2015.8-9.24
    [BibTeX]
    @InProceedings{2015-09-CICHOWSKI,
    author = {Janusz Cichowski and Lisowski Karol and Szczuko Piotr and Andrzej Czy\.zewski},
    title = {{Zdalny zintegrowany modu\l{} nadzoru radiowo-wizyjnego (An integrated module for remote radio and visual monitoring)}},
    booktitle = {{Krajowe Sympozjum Telekomunikacji i Teleinformatyki (KSTiT 2015)}},
    address = {Krak\'{o}w, Poland},
    date = {2015-09-13/2015-09-16},
    doi = {10.15199/59.2015.8-9.24},
    year = {2015}
    }

  • Eftychia Datsika, Angelos Antonopoulos, Nizar Zorba, and Christos Verikoukis. Adaptive Cooperative Network Coding based MAC protocol for Device-to-Device Communication. In IEEE Proceedings of International Conference on Communications (ICC 2015), pages 6996-7001, London, UK, 2015. doi:10.1109/ICC.2015.7249441
    [BibTeX] [Abstract]

    Device-to-Device (D2D) communication allows the direct connection of mobile devices in a cellular network. In D2D networking, cooperative communication is inherent due to the proximity of the devices that are able to overhear and forward information. Particularly, adjacent devices can act as relays and assist the communication of other devices. Network Coding (NC) can further increase the cooperation gains, since a number of packets can be encoded and transmitted together. However, the contention among multiple relays causes channel access issues that must be regulated by effective Medium Access Control (MAC) protocols. In this context, we propose an Adaptive Cooperative Network Coding-based MAC (ACNC-MAC) protocol that utilizes cooperative relaying and exploits NC opportunities in a D2D topology. Both analytical and simulation results show that the proposed protocol is advantageous in terms of energy efficiency without sacrificing the Quality of Service (QoS).

    @InProceedings{2015-06-DATISKA,
    author = {Eftychia Datsika and Angelos Antonopoulos and Nizar Zorba and Christos Verikoukis},
    title = {{Adaptive Cooperative Network Coding based MAC protocol for Device-to-Device Communication}},
    booktitle = {{IEEE Proceedings of International Conference on Communications (ICC 2015)}},
    date = {2015-06-08/2015-06-12},
    address = {London, UK},
    year = {2015},
    pages = {6996-7001},
    doi = {10.1109/ICC.2015.7249441},
    abstract = {Device-to-Device (D2D) communication allows the direct connection of mobile devices in a cellular network. In D2D networking, cooperative communication is inherent due to the proximity of the devices that are able to overhear and forward information. Particularly, adjacent devices can act as relays and assist the communication of other devices. Network Coding (NC) can further increase the cooperation gains, since a number of packets can be encoded and transmitted together. However, the contention among multiple relays causes channel access issues that must be regulated by effective Medium Access Control (MAC) protocols. In this context, we propose an Adaptive Cooperative Network Coding-based MAC (ACNC-MAC) protocol that utilizes cooperative relaying and exploits NC opportunities in a D2D topology. Both analytical and simulation results show that the proposed protocol is advantageous in terms of energy efficiency without sacrificing the Quality of Service (QoS).}
    }

  • Nicklas Bo Jensen, Niklas Quarfot Nielsen, Gregory L. Lee, Sven Karlsson, Matthew LeGendre, Martin Schulz, and Dong H. Ahn. A Scalable Prescriptive Parallel Debugging Model. In Proceedings of IEEE International Parallel and Distributed Processing Symposium, pages 473-483, Hyderabad, India, 2015. doi:10.1109/IPDPS.2015.15
    [BibTeX] [Abstract]

    Debugging is a critical step in the development of any parallel program. However, the traditional interactive debugging model, where users manually step through code and inspect their application, does not scale well even for current supercomputers due its centralized nature. While lightweight debugging models, which have been proposed as an alternative, scale well, they can currently only debug a subset of bug classes. We therefore propose a new model, which we call prescriptive debugging, to fill this gap between these two approaches. This user-guided model allows programmers to express and test their debugging intuition in a way that helps to reduce the error space. Based on this debugging model we introduce a prototype implementation embodying this model, the DySectAPI, allowing programmers to construct probe trees for automatic, event-driven debugging at scale. In this paper we introduce the concepts behind DySectAPI and, using both experimental results and analytical modelling, we show that the DySectAPI implementation can run with a low overhead on current systems. We achieve a logarithmic scaling of the prototype and show predictions that even for a large system the overhead of the prescriptive debugging model will be small.

    @InProceedings{2015-05-JENSEN-1,
    author = {Nicklas Bo Jensen and Niklas Quarfot Nielsen and Gregory L. Lee and Sven Karlsson and Matthew LeGendre and Martin Schulz and Dong H. Ahn},
    title = {{A Scalable Prescriptive Parallel Debugging Model}},
    booktitle = {{Proceedings of IEEE International Parallel and Distributed Processing Symposium}},
    date = {2015-05-25/2015-05-27},
    address = {Hyderabad, India},
    doi = {10.1109/IPDPS.2015.15},
    pages = {473-483},
    abstract = {Debugging is a critical step in the development of any parallel program. However, the traditional interactive debugging model, where users manually step through code and inspect their application, does not scale well even for current supercomputers due its centralized nature. While lightweight debugging models, which have been proposed as an alternative, scale well, they can currently only debug a subset of bug classes. We therefore propose a new model, which we call prescriptive debugging, to fill this gap between these two approaches. This user-guided model allows programmers to express and test their debugging intuition in a way that helps to reduce the error space. Based on this debugging model we introduce a prototype implementation embodying this model, the DySectAPI, allowing programmers to construct probe trees for automatic, event-driven debugging at scale. In this paper we introduce the concepts behind DySectAPI and, using both experimental results and analytical modelling, we show that the DySectAPI implementation can run with a low overhead on current systems. We achieve a logarithmic scaling of the prototype and show predictions that even for a large system the overhead of the prescriptive debugging model will be small.},
    year = {2015}
    }

  • Nicklas Bo Jensen, Pascal Schleuniger, Andreas Hindborg, Maxwell Walter, and Sven Karlsson. Experiences with Compiler Support for Processors with Exposed Pipelines. In Proceedings of the 22nd Reconfigurable Architectures Workshop (RAW 2015), pages 137-143, Hyderabad, India, 2015. doi:10.1109/IPDPSW.2015.9
    [BibTeX] [Abstract]

    Field programmable gate arrays, FPGAs, have become an attractive implementation technology for a broad range of computing systems. We recently proposed a processor architecture, Tinuso, which achieves high performance by moving complexity from hardware to the compiler tool chain. This means that the compiler tool chain must handle the increased complexity. However, it is not clear if current production compilers can successfully meet the strict constraints on instruction order and generate efficient object code. In this paper, we present our experiences developing a compiler backend using the GNU Compiler Collection, GCC. For a set of C benchmarks, we show that a Tinuso implementation with our GCC backend reaches a relative speedup of up to 1.73 over a similar Xilinx Micro Blaze configuration while using 30% fewer hardware resources. While our experiences are generally positive, we expose some limitations in GCC that need to be addressed to achieve the full performance potential of Tinuso.

    @InProceedings{2015-05-JENSEN-2,
    author = {Nicklas Bo Jensen and Pascal Schleuniger and Andreas Hindborg and Maxwell Walter and Sven Karlsson},
    title = {{Experiences with Compiler Support for Processors with Exposed Pipelines}},
    booktitle = {{Proceedings of the 22nd Reconfigurable Architectures Workshop (RAW 2015)}},
    date = {2015-05-25},
    address = {Hyderabad, India},
    doi = {10.1109/IPDPSW.2015.9},
    pages = {137-143},
    abstract = {Field programmable gate arrays, FPGAs, have become an attractive implementation technology for a broad range of computing systems. We recently proposed a processor architecture, Tinuso, which achieves high performance by moving complexity from hardware to the compiler tool chain. This means that the compiler tool chain must handle the increased complexity. However, it is not clear if current production compilers can successfully meet the strict constraints on instruction order and generate efficient object code. In this paper, we present our experiences developing a compiler backend using the GNU Compiler Collection, GCC. For a set of C benchmarks, we show that a Tinuso implementation with our GCC backend reaches a relative speedup of up to 1.73 over a similar Xilinx Micro Blaze configuration while using 30% fewer hardware resources. While our experiences are generally positive, we expose some limitations in GCC that need to be addressed to achieve the full performance potential of Tinuso.},
    year = {2015}
    }

  • Valentin Koblar, Martin Pečar, Klemen Gantar, Tea Tušar, and Bogdan Filipič. Determining surface roughness of semifinished products using computer vision and machine learning. In Proceedings of the 18th International Multiconference Information Society (IS 2015), volume A, pages 51-54, 2015.
    [BibTeX] [Abstract] [Download PDF]

    In the production of components for various industries, including automotive, monitoring of surface roughness is one of the key quality control procedures since achieving appropriate surface quality is necessary for reliable functioning of the manufactured components. This study deals with the task of determining the surface roughness of semifinished products and proposes a computer-vision-based method for this purpose. To automate the design of the method, machine learning is used to induce suitable predictive models from the captured product images, and evolutionary computation to tune the computer vision algorithm parameters. The resulting method allows for accurate online determination of roughness quality classes and shows a potential for online prediction of roughness values.

    @InProceedings{2015-10-KOBLAR,
    title = {{Determining surface roughness of semifinished products using computer vision and machine learning}},
    author = {Valentin Koblar and Martin Pe\v{c}ar and Klemen Gantar and Tea Tu\v{s}ar and Bogdan Filipi\v{c}},
    booktitle = {{Proceedings of the 18th International Multiconference Information Society (IS 2015)}},
    volume = {A},
    pages = {51-54},
    date = {2015-10},
    url = {http://www.copcams.eu/wp-content/uploads/2015/10/Koblar_etal_IS2015_Vol.A_51-54.pdf},
    abstract = {In the production of components for various industries, including automotive, monitoring of surface roughness is one of the key quality control procedures since achieving appropriate surface quality is necessary for reliable functioning of the manufactured components. This study deals with the task of determining the surface roughness of semifinished products and proposes a computer-vision-based method for this purpose. To automate the design of the method, machine learning is used to induce suitable predictive models from the captured product images, and evolutionary computation to tune the computer vision algorithm parameters. The resulting method allows for accurate online determination of roughness quality classes and shows a potential for online prediction of roughness values.},
    year = {2015}
    }

  • Karol Lisowski and Andrzej Czyżewski. Complexity Analysis of the Pawlak’s Flowgraph Extension for Re-Identification in Multi-Camera Surveillance System. Multimedia Tools and Applications, 75(17):10495-10511, 2015. doi:10.1007/s11042-015-2652-z
    [BibTeX] [Abstract]

    The idea of Pawlak’s flowgraph turned out to be a useful and convenient container for a knowledge of objects’ behaviour and movements within the area observed with a multi-camera surveillance system. Utilization of the flowgraph for modelling behaviour admittedly requires certain extensions and enhancements, but it allows for combining many rules into a one data structure and for obtaining parameters describing how objects tend to move through the supervised area. The main aim of this article is presentation of the complexity analysis of proposed modification of flowgraphs. This analysis contains considerations of issues such as memory efficiency and computational complexity of operations on the flowgraph. The measures related to space and time efficiency were also included.

    @Article{2015-05-LISOWSKI,
    author = {Karol Lisowski and Andrzej Czy\.zewski},
    title = {{Complexity Analysis of the Pawlak's Flowgraph Extension for Re-Identification in Multi-Camera Surveillance System}},
    journal = {{Multimedia Tools and Applications}},
    date = {2015-05-26},
    volume = {75},
    number = {17},
    pages = {10495-10511},
    doi = {10.1007/s11042-015-2652-z},
    publisher = {Springer},
    abstract = {The idea of Pawlak's flowgraph turned out to be a useful and convenient container for a knowledge of objects’ behaviour and movements within the area observed with a multi-camera surveillance system. Utilization of the flowgraph for modelling behaviour admittedly requires certain extensions and enhancements, but it allows for combining many rules into a one data structure and for obtaining parameters describing how objects tend to move through the supervised area. The main aim of this article is presentation of the complexity analysis of proposed modification of flowgraphs. This analysis contains considerations of issues such as memory efficiency and computational complexity of operations on the flowgraph. The measures related to space and time efficiency were also included.},
    year = {2015}
    }

  • Karol Lisowski and Andrzej Czyżewski. Pawlak’s Flow Graph Extensions for Video Surveillance Systems. In Proceedings of the 10th International Symposium Advances in Artificial Intelligence and Applications (AAIA 2015), pages 81-87, Lodz, Poland, 2015. doi:10.15439/2015F384
    [BibTeX]

  • Prodromos-Vasileios Mekikis, Elli Kartsakli, Aris S. Lalos, Angelos Antonopoulos, Luis Alonso, and Christos Verikoukis. Connectivity of Large-Scale WSNs in Fading Environments under Different Routing Mechanisms. In Proceedings of IEEE International Conference on Communications (ICC 2015), pages 6553-6558, London, UK, 2015. doi:10.1109/ICC.2015.7249369
    [BibTeX] [Abstract]

    As the number of nodes in wireless sensor networks (WSNs) increases, new challenges have to be faced in order to maintain their performance. A fundamental requirement of several applications is the correct transmission of the measurements to their final destinations. Thus, it is crucial to guarantee a high probability of connectivity, which characterizes the ability of every node to report to the fusion center. This network metric is strongly affected by both the fading characteristics and the different routing protocols that are used for the dissemination of data. In this paper, we study the probability of a network to be fully connected for two widely employed routing mechanisms, namely unicast and K-anycast. The analytical derivations and the simulations evaluate the trade-offs among the different routing mechanisms and provide useful guidelines on the design of WSNs.

    @InProceedings{2015-06-MEKIKIS,
    author = {Prodromos-Vasileios Mekikis and Elli Kartsakli and Aris S. Lalos and Angelos Antonopoulos and Luis Alonso and Christos Verikoukis},
    title = {{Connectivity of Large-Scale WSNs in Fading Environments under Different Routing Mechanisms}},
    booktitle = {{Proceedings of IEEE International Conference on Communications (ICC 2015)}},
    date = {2015-06-08/2015-06-12},
    year = {2015},
    pages = {6553-6558},
    doi = {10.1109/ICC.2015.7249369},
    address = {London, UK},
    abstract = {As the number of nodes in wireless sensor networks (WSNs) increases, new challenges have to be faced in order to maintain their performance. A fundamental requirement of several applications is the correct transmission of the measurements to their final destinations. Thus, it is crucial to guarantee a high probability of connectivity, which characterizes the ability of every node to report to the fusion center. This network metric is strongly affected by both the fading characteristics and the different routing protocols that are used for the dissemination of data. In this paper, we study the probability of a network to be fully connected for two widely employed routing mechanisms, namely unicast and K-anycast. The analytical derivations and the simulations evaluate the trade-offs among the different routing mechanisms and provide useful guidelines on the design of WSNs.}
    }

  • Anca Molnos, Warody Lombardi, Suzanne Lesecq, Julien Mottin, Diego Puschini, and Arnaud Tonda. Energy Management via PI Control for Data Parallel Applications with Throughput Constraints. In Proceedings of the 25th International Workshop on Power and Timing Modeling, Optimization and Simulation (PATMOS 2015), pages 63-70, Salvador da Bahia, Brazil, 2015. doi:10.1109/PATMOS.2015.7347588
    [BibTeX]
    @InProceedings{2015-09-MOLNOS,
    author = {Anca Molnos and Warody Lombardi and Suzanne Lesecq and Julien Mottin and Diego Puschini and Arnaud Tonda},
    title = {{Energy Management via PI Control for Data Parallel Applications with Throughput Constraints}},
    booktitle = {{Proceedings of the 25th International Workshop on Power and Timing Modeling, Optimization and Simulation (PATMOS 2015)}},
    address = {Salvador da Bahia, Brazil},
    date = {2015-09-01/2015-09-04},
    doi = {10.1109/PATMOS.2015.7347588},
    pages = {63-70},
    year = {2015}
    }

  • Jesper Puge Nielsen, Maxwell Walter, and Sven Karlsson. Designing a Lock-free Virtual Memory Management System. In Proceedings of the Swedish Workshop on Multicore Computing (MCC 2015), Copenhagen, DK, 2015.
    [BibTeX] [Abstract]

    Operating system developers must turn to new techniques to maintain scalability as the number of cores controlled by the operating system grows. Traditional lock based system design has proven to be easy to implement, but their performance fails to scale with the number of cores. In this paper we present the design of a virtual memory management system for a multi-core operating system. The system is designed without the use of locks. Instead, atomic operations are used to synchronize between threads. We discuss the challenges of designing a lock-free virtual memory management system, and present the solutions we have chosen. These includes a flat virtual address space design supporting lock-free concurrent modifications, and a new algorithm for sharing physical page frames between processes, without using locks to synchronize new physical page frame mappings and concurrent reclamation of physical page frames. We have implemented our design for the AMD64 architecture and evaluated it by comparing it to a lock based implementation.

    @InProceedings{2015-11-NIELSEN,
    title = {{Designing a Lock-free Virtual Memory Management System}},
    author = {Jesper Puge Nielsen and Maxwell Walter and Sven Karlsson},
    booktitle = {{Proceedings of the Swedish Workshop on Multicore Computing (MCC 2015)}},
    address = {Copenhagen, DK},
    date = {2015-11-25/2015-11-26},
    year = {2015},
    abstract = {Operating system developers must turn to new techniques to maintain scalability as the number of cores controlled by the operating system grows. Traditional lock based system design has proven to be easy to implement, but their performance fails to scale with the number of cores.
    In this paper we present the design of a virtual memory management system for a multi-core operating system. The system is designed without the use of locks. Instead, atomic operations are used to synchronize between threads. We discuss the challenges of designing a lock-free virtual memory management system, and present the solutions we have chosen. These includes a flat virtual address space design supporting lock-free concurrent modifications, and a new algorithm for sharing physical page frames between processes, without using locks to synchronize new physical page frame mappings and concurrent reclamation of physical page frames. We have implemented our design for the AMD64 architecture and evaluated it by comparing it to a lock based implementation.}
    }

  • Fabio Poiesi and Andrea Cavallaro. MTTV – An Interactive Trajectory Visualization and Analysis Tool. In Proceedings of Information Visualization Theory and Applications (IVAPP), pages 157-162, Berlin, Germany, 2015.
    [BibTeX] [Abstract]

    We present an interactive visualizer that enables the exploration, measurement, analysis and manipulation of trajectories. Trajectories can be generated either automatically by multi-target tracking algorithms or manually by human annotators. The visualizer helps understanding the behavior of targets, correcting tracking results and quantifying the performance of tracking algorithms. The input video can be overlaid to compare ideal and estimated target locations. The code of the visualizer (C++ with openFrameworks) is open source.

    @InProceedings{2015-03-POIESI,
    author = {Fabio Poiesi and Andrea Cavallaro},
    title = {{MTTV - An Interactive Trajectory Visualization and Analysis Tool}},
    booktitle = {{Proceedings of Information Visualization Theory and Applications (IVAPP)}},
    date = {2015-03-11/2015-03-14},
    address = {Berlin, Germany},
    pages = {157-162},
    abstract = {We present an interactive visualizer that enables the exploration, measurement, analysis and manipulation of trajectories. Trajectories can be generated either automatically by multi-target tracking algorithms or manually by human annotators. The visualizer helps understanding the behavior of targets, correcting tracking results and quantifying the performance of tracking algorithms. The input video can be overlaid to compare ideal and estimated target locations. The code of the visualizer (C++ with openFrameworks) is open source.},
    year = {2015}
    }

  • Fabio Poiesi and Andrea Cavallaro. Self-positioning of a team of flying smart cameras. In IEEE Proceedings of Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP 2015), pages 1-6, Singapore, 2015. doi:10.1109/ISSNIP.2015.7106943
    [BibTeX] [Abstract]

    Quadcopters are highly maneuverable and can provide an effective means for an agile dynamic positioning of sensors such as cameras. In this paper we propose a method for the self-positioning of a team of camera-equipped quadcopters (flying cameras) around a moving target. The self-positioning task is driven by the maximization of the monitored surface of the moving target based on a dynamic flight model combined with a collision avoidance algorithm. Each flying camera only knows the relative distance of neighboring flying cameras and its desired position with respect to the target. Given a team of up to 12 flying cameras, we show they can achieve a stable time-varying formation around a moving target without collisions.

    @InProceedings{2015-04-POIESI,
    title = {{Self-positioning of a team of flying smart cameras}},
    author = {Fabio Poiesi and Andrea Cavallaro},
    booktitle = {{IEEE Proceedings of Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP 2015)}},
    address = {Singapore},
    date = {2015-04-07/2015-04-09},
    doi = {10.1109/ISSNIP.2015.7106943},
    year = {2015},
    pages = {1-6},
    abstract = {Quadcopters are highly maneuverable and can provide an effective means for an agile dynamic positioning of sensors such as cameras. In this paper we propose a method for the self-positioning of a team of camera-equipped quadcopters (flying cameras) around a moving target. The self-positioning task is driven by the maximization of the monitored surface of the moving target based on a dynamic flight model combined with a collision avoidance algorithm. Each flying camera only knows the relative distance of neighboring flying cameras and its desired position with respect to the target. Given a team of up to 12 flying cameras, we show they can achieve a stable time-varying formation around a moving target without collisions.}
    }

  • Fabio Poiesi and Andrea Cavallaro. Distributed Vision-Based Flying Cameras to Film a Moving Target. In Proceedings of International Conference on Intelligent Robots and Systems (IROS 2015), pages 2453-2459, Hambourg, Germany, 2015. doi:10.1109/IROS.2015.7353710
    [BibTeX] [Abstract] [Download PDF]

    Formations of camera-equipped quadrotors (flying cameras) have the actuation agility to track moving targets from multiple viewing angles. In this paper we propose an infrastructure-free distributed control method for multiple flying cameras tracking a moving object. The proposed visionbased servoing can deal with noisy and missing target observations, accounts for quadrotor oscillations and does not require an external positioning system. The flight direction of each camera is inferred via geometric derivation, and the formation is maintained by employing a distributed algorithm that uses the target position information on the camera plane and the position of neighboring flying cameras. Simulations show that the proposed solution enables the tracking of a moving target by the cameras flying in formation despite noisy target detections and when the target is outside some of the fields of view.

    @InProceedings{2015-09-POIESI,
    author = {Fabio Poiesi and Andrea Cavallaro},
    title = {{Distributed Vision-Based Flying Cameras to Film a Moving Target}},
    booktitle = {{Proceedings of International Conference on Intelligent Robots and Systems (IROS 2015)}},
    address = {Hambourg, Germany},
    date = {2015-09-28/2015-10-02},
    url = {http://www.fabio-poiesi.com/files/papers/conferences/2015_IROS_DistributedFlyingCamerasFilmTarget_Poiesi_Cavallaro.pdf},
    abstract = {Formations of camera-equipped quadrotors (flying cameras) have the actuation agility to track moving targets from multiple viewing angles. In this paper we propose an infrastructure-free distributed control method for multiple flying cameras tracking a moving object. The proposed visionbased servoing can deal with noisy and missing target observations, accounts for quadrotor oscillations and does not require an external positioning system. The flight direction of each camera is inferred via geometric derivation, and the formation is maintained by employing a distributed algorithm that uses the target position information on the camera plane and the position of neighboring flying cameras. Simulations show that the proposed solution enables the tracking of a moving target by the cameras flying in formation despite noisy target detections and when the target is outside some of the fields of view.},
    year = {2015},
    doi = {10.1109/IROS.2015.7353710},
    pages = {2453-2459}
    }

  • Tiana Rakotovao, Julien Mottin, Diego Puschini, and Christian Laugier. Real-Time Power-Efficient Integration of Multi-Sensor Occupancy Grid on Many-Core. In The 2015 IEEE International Workshop on Advanced Robotics and its Social Impacts (ARSO 2015), pages 1-6, Lyon, France, 2015. doi:10.1109/ARSO.2015.7428211
    [BibTeX]

  • Michal Tarkowski, Przemysł, and Lukasz Kulas. Efficient Algorithm for Blinking LED Detection Dedicated to Embedded Systems Equipped with High Performance Cameras. In Proceedings of International Conference on Computer as a Tool (EUROCON 2015), pages 1-4, Salamanca, Spain, 2015. doi:10.1109/EUROCON.2015.7313723
    [BibTeX]
    @InProceedings{2015-09-TARKOWSKI,
    author = {Michal Tarkowski and Przemys\l{}aw Wo\'{z}nica and \Lukasz Kulas},
    booktitle = {{Proceedings of International Conference on Computer as a Tool (EUROCON 2015)}},
    title = {{Efficient Algorithm for Blinking LED Detection Dedicated to Embedded Systems Equipped with High Performance Cameras}},
    date = {2015-09-08/2015-09-11},
    address = {Salamanca, Spain},
    doi = {10.1109/EUROCON.2015.7313723},
    pages = {1-4},
    year = {2015}
    }

  • Tea Tušar and Bogdan Filipič. Visualization of Pareto front approximations in evolutionary multiobjective optimization: A critical review and the prosection method. IEEE Transactions on Evolutionary Computation, 19(2):225-245, 2015. doi:10.1109/TEVC.2014.2313407
    [BibTeX] [Abstract]

    In evolutionary multiobjective optimization, it is very important to be able to visualize approximations of the Pareto front (called approximation sets) that are found by multiobjective evolutionary algorithms. While scatter plots can be used for visualizing 2-D and 3-D approximation sets, more advanced approaches are needed to handle four or more objectives. This paper presents a comprehensive review of the existing visualization methods used in evolutionary multiobjective optimization, showing their outcomes on two novel 4-D benchmark approximation sets. In addition, a visualization method that uses prosection (projection of a section) to visualize 4-D approximation sets is proposed. The method reproduces the shape, range, and distribution of vectors in the observed approximation sets well and can handle multiple large approximation sets while being robust and computationally inexpensive. Even more importantly, for some vectors, the visualization with prosections preserves the Pareto dominance relation and relative closeness to reference points. The method is analyzed theoretically and demonstrated on several approximation sets.

    @article{2015-03-TUSAR,
    title = {{Visualization of Pareto front approximations in evolutionary multiobjective optimization: A critical review and the prosection method}},
    author = {Tea Tu\v{s}ar and Bogdan Filipi\v{c}},
    journal = {{IEEE Transactions on Evolutionary Computation}},
    volume = {19},
    number = {2},
    pages = {225-245},
    year = {2015},
    date = {2015-04},
    abstract = {In evolutionary multiobjective optimization, it is very important to be able to visualize approximations of the Pareto front (called approximation sets) that are found by multiobjective evolutionary algorithms. While scatter plots can be used for visualizing 2-D and 3-D approximation sets, more advanced approaches are needed to handle four or more objectives. This paper presents a comprehensive review of the existing visualization methods used in evolutionary multiobjective optimization, showing their outcomes on two novel 4-D benchmark approximation sets. In addition, a visualization method that uses prosection (projection of a section) to visualize 4-D approximation sets is proposed. The method reproduces the shape, range, and distribution of vectors in the observed approximation sets well and can handle multiple large approximation sets while being robust and computationally inexpensive. Even more importantly, for some vectors, the visualization with prosections preserves the Pareto dominance relation and relative closeness to reference points. The method is analyzed theoretically and demonstrated on several approximation sets.},
    doi = {10.1109/TEVC.2014.2313407}
    }

  • Sven Verdoolaege, Tobias Grosser, and Albert Cohen. Polyhedral AST Generation is More than Scanning Polyhedra. Acm transactions on programming languages and systems, 37(4):12:1-12:50, 2015. doi:10.1145/2743016
    [BibTeX] [Abstract]

    Abstract mathematical representations such as integer polyhedra have been shown to be useful to precisely analyze computational kernels and to express complex loop transformations. Such transformations rely on abstract syntax tree (AST) generators to convert the mathematical representation back to an imperative program. Such generic AST generators avoid the need to resort to transformation-specific code generators, which may be very costly or technically difficult to develop as transformations become more complex. Existing AST generators have proven their effectiveness, but they hit limitations in more complex scenarios. Specifically, (1) they do not support or may fail to generate control flow for complex transformations using piecewise schedules or mappings involving modulo arithmetic; (2) they offer limited support for the specialization of the generated code exposing compact, straightline, vectorizable kernels with high arithmetic intensity necessary to exploit the peak performance of modern hardware; (3) they offer no support for memory layout transformations; and (4) they provide insufficient control over the AST generation strategy, preventing their application to complex domain-specific optimizations. We present a new AST generation approach that extends classical polyhedral scanning to the full generality of Presburger arithmetic, including existentially quantified variables and piecewise schedules, and introduce new optimizations for the detection of components and shifted strides. Not limiting ourselves to control flow generation, we expose functionality to generate AST expressions from arbitrary piecewise quasi-affine expressions, which enables the use of our AST generator for data-layout transformations. We complement this with support for specialization by polyhedral unrolling, user-directed versioning, and specialization of AST expressions according to the location at which they are generated, and we complete this work with fine-grained user control over the AST generation strategies used. Using this generalized idea of AST generation, we present how to implement complex domain-specific transformations without the need to write specialized code generators, but instead relying on a generic AST generator parametrized to a specific problem domain.

    @Article{2015-08-VERDOOLAEGE,
    author = {Sven Verdoolaege and Tobias Grosser and Albert Cohen},
    title = {{Polyhedral {AST} Generation is More than Scanning Polyhedra}},
    journal = {ACM Transactions on Programming Languages and Systems},
    date = {2015-08},
    volume = {37},
    number = {4},
    year = {2015},
    pages = {12:1-12:50},
    doi = {10.1145/2743016},
    publisher = {ACM},
    address = {New York, NY, USA},
    abstract = {Abstract mathematical representations such as integer polyhedra have been shown to be useful to precisely analyze computational kernels and to express complex loop transformations. Such transformations rely on abstract syntax tree (AST) generators to convert the mathematical representation back to an imperative program. Such generic AST generators avoid the need to resort to transformation-specific code generators, which may be very costly or technically difficult to develop as transformations become more complex. Existing AST generators have proven their effectiveness, but they hit limitations in more complex scenarios. Specifically, (1) they do not support or may fail to generate control flow for complex transformations using piecewise schedules or mappings involving modulo arithmetic; (2) they offer limited support for the specialization of the generated code exposing compact, straightline, vectorizable kernels with high arithmetic intensity necessary to exploit the peak performance of modern hardware; (3) they offer no support for memory layout transformations; and (4) they provide insufficient control over the AST generation strategy, preventing their application to complex domain-specific optimizations. We present a new AST generation approach that extends classical polyhedral scanning to the full generality of Presburger arithmetic, including existentially quantified variables and piecewise schedules, and introduce new optimizations for the detection of components and shifted strides. Not limiting ourselves to control flow generation, we expose functionality to generate AST expressions from arbitrary piecewise quasi-affine expressions, which enables the use of our AST generator for data-layout transformations. We complement this with support for specialization by polyhedral unrolling, user-directed versioning, and specialization of AST expressions according to the location at which they are generated, and we complete this work with fine-grained user control over the AST generation strategies used. Using this generalized idea of AST generation, we present how to implement complex domain-specific transformations without the need to write specialized code generators, but instead relying on a generic AST generator parametrized to a specific problem domain.}
    }

  • Maxwell Walter, Jesper Puge Nielsen, and Sven Karlsson. Efficient Communication Using Ring Buffers — Early Results. In Proceedings of the Swedish Workshop on Multicore Computing (MCC 2015), Copenhagen, DK, 2015.
    [BibTeX] [Abstract]

    Communication is a vital concern to modern multiprocessor and multi-core systems, as communication between threads and processes on different cores can quickly become a performance bottle-neck. In addition, most communication mechanisms require synchronization protocols that are complicated to get right, further degrading performance. We propose a communication mechanism, based on shared memory, with built-in synchronization protocols. We model our communication structures, which we call ring buffers, around the communication mechanisms observed between current hardware and software systems. While our ring buffers are not operating system specific, we have designed our multi-core research operating system, FenixOS, around them. This makes communication a fundamental system component and allows us to provide efficient communication between threads, processes, the kernel, and device drivers. In this paper we describe our preliminary work on our ring buffer structures. We also provide early results comparing our ring buffers against several other common communication mechanisms, as well as a Linux ring buffer implementation.

    @InProceedings{2015-11-WALTER,
    title = {{Efficient Communication Using Ring Buffers -- Early Results}},
    author = {Maxwell Walter and Jesper Puge Nielsen and Sven Karlsson},
    booktitle = {{Proceedings of the Swedish Workshop on Multicore Computing (MCC 2015)}},
    address = {Copenhagen, DK},
    date = {2015-11-25/2015-11-26},
    year = {2015},
    abstract = {Communication is a vital concern to modern multiprocessor and multi-core systems, as communication between threads and processes on different cores can quickly become a performance bottle-neck. In addition, most communication mechanisms require synchronization protocols that are complicated to get right, further degrading performance. We propose a communication mechanism, based on shared memory, with built-in synchronization protocols. We model our communication structures, which we call ring buffers, around the communication mechanisms observed between current hardware and software systems.
    While our ring buffers are not operating system specific, we have designed our multi-core research operating system, FenixOS, around them. This makes communication a fundamental system component and allows us to provide efficient communication between threads, processes, the kernel, and device drivers. In this paper we describe our preliminary work on our ring buffer structures. We also provide early results comparing our ring buffers against several other common communication mechanisms, as well as a Linux ring buffer implementation.}
    }

2014

  • Toygar Akgün, Charles Attwood, Andrea Cavallaro, Christian Fabre, Fabio Poiesi, and Piotr Szczuko. Human Behavior Understanding in Networked Sensing, chapter Towards Cognitive and Perceptive Video Systems, pages 3-17. Springer, 2014. doi:10.1007/978-3-319-10807-0_1
    [BibTeX] [Abstract]

    In this chapter we cover research and development issues related to smart cameras. We discuss challenges, new technologies and algorithms, applications and the evaluation of today’s technologies. We will cover problems related to software, hardware, communication, embedded and distributed systems, multi-modal sensors, privacy and security. We also discuss future trends and market expectations from the customer’s point of view.

    @InBook{2014-11-AKGUN,
    author = {Toygar Akg\"{u}n and Charles Attwood and Andrea Cavallaro and Christian Fabre and Fabio Poiesi and Piotr Szczuko},
    chapter = {Towards Cognitive and Perceptive Video Systems},
    editor = {Paolo Spagnolo and Pier Luigi Mazzeo and Cosimo Distante},
    title = {{Human Behavior Understanding in Networked Sensing}},
    pages = {3-17},
    date = {2014-11-07},
    publisher = {Springer},
    doi = {10.1007/978-3-319-10807-0_1},
    abstract = {In this chapter we cover research and development issues related to smart cameras. We discuss challenges, new technologies and algorithms, applications and the evaluation of today's technologies. We will cover problems related to software, hardware, communication, embedded and distributed systems, multi-modal sensors, privacy and security. We also discuss future trends and market expectations from the customer's point of view.},
    year = {2014}
    }

  • Angelos Antonopoulos and Christos Verikoukis. Multi-Player Game Theoretic MAC Strategies for Energy Efficient Data Dissemination. IEEE Transactions on Wireless Communications, 13(2):592-603, 2014. doi:10.1109/TWC.2013.120713.120790
    [BibTeX] [Abstract]

    The existence of multiple active sources in wireless data dissemination scenarios raises considerable channel access issues. Although the overall goal in the network is the successful accomplishment of the dissemination, the individual wireless nodes aim at maximizing their battery lifetime by minimizing their particular energy consumption. Considering thus the self-centered behavior of the nodes, we anticipate to design Medium Access Control (MAC) policies in order to provide the terminals with energy efficient solutions without compromising the dissemination completion time. In this paper, we propose non-cooperative game theoretic channel access strategies by estimating equilibrium points that achieve balance between conserving energy and completing the data dissemination. In particular, we present two different MAC strategies: i) a distributed approach for ad-hoc networks, and ii) a coordinated approach for infrastructure networks, where a central controller is sporadically used to accelerate the data dissemination. In addition, network-coded transmissions are considered to eliminate the need of control packets exchange. Both analytical and simulation results are provided to evaluate our proposed schemes, demonstrating the significant gains that game theoretic techniques can bring to the network performance compared to standardized solutions.

    @Article{2014-02-ANTONOPOULOS,
    author = {Angelos Antonopoulos and Christos Verikoukis},
    journal = {{IEEE Transactions on Wireless Communications}},
    title = {{Multi-Player Game Theoretic MAC Strategies for Energy Efficient Data Dissemination}},
    date = {2014-02-11},
    volume = {13},
    number = {2},
    pages = {592-603},
    doi = {10.1109/TWC.2013.120713.120790},
    abstract = {The existence of multiple active sources in wireless data dissemination scenarios raises considerable channel access issues. Although the overall goal in the network is the successful accomplishment of the dissemination, the individual wireless nodes aim at maximizing their battery lifetime by minimizing their particular energy consumption. Considering thus the self-centered behavior of the nodes, we anticipate to design Medium Access Control (MAC) policies in order to provide the terminals with energy efficient solutions without compromising the dissemination completion time. In this paper, we propose non-cooperative game theoretic channel access strategies by estimating equilibrium points that achieve balance between conserving energy and completing the data dissemination. In particular, we present two different MAC strategies: i) a distributed approach for ad-hoc networks, and ii) a coordinated approach for infrastructure networks, where a central controller is sporadically used to accelerate the data dissemination. In addition, network-coded transmissions are considered to eliminate the need of control packets exchange. Both analytical and simulation results are provided to evaluate our proposed schemes, demonstrating the significant gains that game theoretic techniques can bring to the network performance compared to standardized solutions.},
    year = {2014}
    }

  • Angelos Antonopoulos, Elli Kartsakli, and Christos Verikoukis. Game Theoretic D2D Content Dissemination in 4G Cellular Networks. IEEE Communications Magazine, 52(6), 2014. doi:10.1109/MCOM.2014.6829954
    [BibTeX]

  • Angelos Antonopoulos, Joao Bastos, and Christos Verikoukis. Analogue Network Coding-aided Game Theoretic Medium Access Control Protocol for Energy-Efficient Data Dissemination. IET Science, Measurement & Technology, 8(6):399-407, 2014. doi:10.1049/iet-smt.2013.0192
    [BibTeX] [Download PDF]
    @Article{2014-11-ANTONOPOULOS,
    author = {Angelos Antonopoulos and Joao Bastos and Christos Verikoukis},
    title = {{Analogue Network Coding-aided Game Theoretic Medium Access Control Protocol for Energy-Efficient Data Dissemination}},
    journal = {{IET Science, Measurement \& Technology}},
    date = {2014-11},
    volume = {8},
    number = {6},
    pages = {399-407},
    url = {http://digital-library.theiet.org/content/journals/iet-smt/8/6},
    year = {2014},
    doi = {10.1049/iet-smt.2013.0192}
    }

  • Laust Brock-Nannestad and Sven Karlsson. Library Support for Resource Constrained Accelerators. In Luiz DeRose, Bronis R. de Supinski, Stephen L. Olivier, Barbara M. Chapman, and Matthias S. Müller, editors, Using and Improving OpenMP for Devices, Tasks, and More, volume 8766 of Lecture Notes in Computer Science, pages 187-201. IEEE, 2014. doi:10.1007/978-3-319-11454-5_14
    [BibTeX] [Abstract]

    Accelerators, and other resource constrained systems, are increasingly being used in computer systems. Accelerators provide power efficient performance and often provide a shared memory model. However, it is a challenge to map feature rich APIs, such as OpenMP, to resource constrained systems. In this paper, we present a lightweight system where an accelerator can remotely execute library functions on a host processor. The implementation takes up 750 bytes but can replace arbitrary library calls leading to significant savings in memory foot print. We evaluate with a set of SPLASH-2 applications and show that the impact on execution time is negligible when compared to GCCs OpenMP implementation.

    @incollection{2014-09-BROCK-NANNESTAD,
    author = {Laust Brock-Nannestad and Sven Karlsson},
    title = {{Library Support for Resource Constrained Accelerators}},
    booktitle = {{Using and Improving OpenMP for Devices, Tasks, and More}},
    series = {Lecture Notes in Computer Science},
    volume = {8766},
    editor = {Luiz DeRose and Bronis R. de Supinski and Stephen L. Olivier and Barbara M. Chapman and Matthias S. M{\"u}ller},
    date = {2014-09-28/2014-09-30},
    pages = {187-201},
    publisher = {IEEE},
    doi = {10.1007/978-3-319-11454-5_14},
    abstract = {Accelerators, and other resource constrained systems, are increasingly being used in computer systems. Accelerators provide power efficient performance and often provide a shared memory model. However, it is a challenge to map feature rich APIs, such as OpenMP, to resource constrained systems. In this paper, we present a lightweight system where an accelerator can remotely execute library functions on a host processor. The implementation takes up 750 bytes but can replace arbitrary library calls leading to significant savings in memory foot print. We evaluate with a set of SPLASH-2 applications and show that the impact on execution time is negligible when compared to GCCs OpenMP implementation.},
    year = {2014}
    }

  • Duygu Büyükaydin, Toygar Akgün, Metin Aktas, and Cengiz Erbas. Video Isleme Algoritmalarinin OpenCL Kullanarak GPU ve FPGA Üzerinde Veri-Paralel Gerçeklenmesi. In Savunma teknolojileri kongresi (SAVTEK 2014), Ankara, Turkey, 2014.
    [BibTeX]
    @InProceedings{2014-06-BUYUKAYADIN,
    author = {Duygu B{\"u}y{\"u}kaydin and Toygar Akg{\"u}n and Metin Aktas and Cengiz Erbas},
    title = {{Video Isleme Algoritmalarinin OpenCL Kullanarak GPU ve FPGA {\"U}zerinde Veri-Paralel Ger{\c c}eklenmesi}},
    booktitle = {Savunma Teknolojileri Kongresi ({SAVTEK 2014})},
    address = {Ankara, Turkey},
    date = {2014-06-25/2014-06-27},
    year = {2014}
    }

  • Janusz Cichowski and Andrzej Czyzewski. Aktywny system RFID do lokalizacji i identyfikacji obiektów w wielomodalnej infrastrukturze bezpieczeństwa (Active RFID System for Objects Localization and Identification in Multimodal Surveillance Infrastructure). Przeglad Telekomunikacyjny – Wiadomosci Telekomunikacyjne (Journal of Telecommunications and Information Technology), 2014(8-9), 2014.
    [BibTeX] [Download PDF]
    @Article{2014-07-CICHOWSKI,
    author = {Janusz Cichowski and Andrzej Czyzewski},
    title = {{Aktywny system RFID do lokalizacji i identyfikacji obiekt{\'o}w w wielomodalnej infrastrukturze bezpiecze{\'n}stwa (Active RFID System for Objects Localization and Identification in Multimodal Surveillance Infrastructure)}},
    journal = {{Przeglad Telekomunikacyjny - Wiadomosci Telekomunikacyjne (Journal of Telecommunications and Information Technology)}},
    date = {2014-07},
    volume = {2014},
    number = {8-9},
    url = {http://www.sigma-not.pl/zeszyt-4182-przeglad-telekomunikacyjny-2014-8-9.html},
    year = {2014}
    }

  • Marco Del Coco and Andrea Cavallaro. Parallel particle-PHD filter. In Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing, Florence, Italy, 2014. doi:10.1109/ICASSP.2014.6854872
    [BibTeX] [Abstract]

    The complexity of multi-target tracking grows faster than linearly with the increase of the numbers of objects, thus making the design of real-time trackers a challenging task for scenarios with a large number of targets. The Probability Hypothesis Density (PHD) filter is known to help reducing this complexity. However, this reduction may not suffice in critical situations when the number of targets, dimension of the state vector, clutter conditions and sample rate are high. To address this problem, we propose a parallelization scheme for the particle PHD filter. The proposed scheme exploits the knowledge of mutual interacting targets in the scene to help fragmentation and to reduce the workload of individual processors. We compare the proposed approach with alternative parallelization schemes and discuss its advantages and limitations using the results obtained on two multi-target tracking datasets.

    @InProceedings{2014-05-COCO,
    title = {{Parallel particle-PHD filter}},
    author = {Marco Del Coco and Andrea Cavallaro},
    booktitle = {{Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing}},
    address= {Florence, Italy},
    date = {2014-05-04/2014-05-09},
    year = {2014},
    doi = {10.1109/ICASSP.2014.6854872},
    abstract = {The complexity of multi-target tracking grows faster than linearly with the increase of the numbers of objects, thus making the design of real-time trackers a challenging task for scenarios with a large number of targets. The Probability Hypothesis Density (PHD) filter is known to help reducing this complexity. However, this reduction may not suffice in critical situations when the number of targets, dimension of the state vector, clutter conditions and sample rate are high. To address this problem, we propose a parallelization scheme for the particle PHD filter. The proposed scheme exploits the knowledge of mutual interacting targets in the scene to help fragmentation and to reduce the workload of individual processors. We compare the proposed approach with alternative parallelization schemes and discuss its advantages and limitations using the results obtained on two multi-target tracking datasets.}
    }

  • Luis Diaz, Eduardo González, Eugenio Villar, and Pablo Sánchez. VIPPE: Native Simulation and Performance Analysis Framework for Multi-Processing Embedded Systems. In Proceedings of the JCE-Sarteco 2014, pages 1-7, Valladolid, Spain, 2014.
    [BibTeX] [Download PDF]
    @InProceedings{2014-09-DIAZ,
    author = {Luis Diaz and Eduardo Gonz\'{a}lez and Eugenio Villar and Pablo S\'{a}nchez},
    title = {{VIPPE: Native Simulation and Performance Analysis Framework for Multi-Processing Embedded Systems}},
    booktitle = {{Proceedings of the JCE-Sarteco 2014}},
    pages = {1-7},
    date = {2014-09-17/2014-09-19},
    address = {Valladolid, Spain},
    url = {http://www.researchgate.net/publication/267265900_VIPPE_Native_simulation_and_performance_analysis_framework_for_multi-processing_embedded_systems},
    year = {2014}
    }

  • Erik Dovgan, Klemen Gantar, Valentin Koblar, and Bogdan Filipič. Detection of Irregularities on Automotive Semiproducts. In Proceedings of the 17th International Multiconference Information Society (IS 2014), pages 22-25, Ljubljana, Slovenia, 2014.
    [BibTeX] [Download PDF]

  • Tobias Grosser, Albert Cohen, Justin Holewinski, P. Sadayappan, and Sven Verdoolaege. Hybrid Hexagonal/Classical Tiling for GPUs. In 12th Annual IEEE/ACM International Symposium on Code Generation and Optimization (CGO ’14), page 66, Orlando, Florida, USA, 2014. doi:10.1145/2544137.2544160
    [BibTeX] [Abstract]

    Time-tiling is necessary for the efficient execution of iterative stencil computations. Classical hyper-rectangular tiles cannot be used due to the combination of backward and forward dependences along space dimensions. Existing techniques trade temporal data reuse for inefficiencies in other areas, such as load imbalance, redundant computations, or increased control flow overhead, therefore making it challenging for use with GPUs. We propose a time-tiling method for iterative stencil computations on GPUs. Our method does not involve redundant computations. It favors coalesced global-memory accesses, data reuse in local/shared-memory or cache, avoidance of thread divergence, and concurrency, combining hexagonal tile shapes along the time and one spatial dimension with classical tiling along the other spatial dimensions. Hexagonal tiles expose multi-level parallelism as well as data reuse. Experimental results demonstrate significant performance improvements over existing stencil compilers.

    @InProceedings{2014-02-GROSSER,
    author = {Tobias Grosser and Albert Cohen and Justin Holewinski and P. Sadayappan and Sven Verdoolaege},
    title = {{Hybrid Hexagonal/Classical Tiling for GPUs}},
    booktitle = {{12th Annual IEEE/ACM International Symposium on Code Generation and Optimization (CGO '14)}},
    date = {2014-02-15/2014-02-19},
    address = {Orlando, Florida, USA},
    pages = {66},
    doi = {10.1145/2544137.2544160},
    abstract = {Time-tiling is necessary for the efficient execution of iterative stencil computations. Classical hyper-rectangular tiles cannot be used due to the combination of backward and forward dependences along space dimensions. Existing techniques trade temporal data reuse for inefficiencies in other areas, such as load imbalance, redundant computations, or increased control flow overhead, therefore making it challenging for use with GPUs. We propose a time-tiling method for iterative stencil computations on GPUs. Our method does not involve redundant computations. It favors coalesced global-memory accesses, data reuse in local/shared-memory or cache, avoidance of thread divergence, and concurrency, combining hexagonal tile shapes along the time and one spatial dimension with classical tiling along the other spatial dimensions. Hexagonal tiles expose multi-level parallelism as well as data reuse. Experimental results demonstrate significant performance improvements over existing stencil compilers.},
    year = {2014}
    }

  • Tobias Grosser, Sven Verdoolaege, Albert Cohen, and (Saday) P. Sadayappan. The relation between diamond tiling and hexagonal tiling. Parallel processing letters, 24(3), 2014. doi:10.1142/S0129626414410023
    [BibTeX]
    @Article{2014-09-GROSSER,
    author = {Tobias Grosser and Sven Verdoolaege and Albert Cohen and P. (Saday) Sadayappan},
    title = {The Relation Between Diamond Tiling and Hexagonal Tiling},
    journal = {Parallel Processing Letters},
    date = {2014-09},
    year = {2014},
    volume = 24,
    number = 3,
    doi = {10.1142/S0129626414410023},
    publisher = {World Scientific Publishing Co.}
    }

  • Nicolas Hili, Christian Fabre, Ivan Llopard, Sophie Dupuy-Chessa, and Dominique Rieu. Model-Based Platform Composition for Embedded System Design. In Proceedings of IEEE 8th International Symposium on Embedded Multicore Many-core Systems-on-Chip (MCSoC-14), pages 157-164, University of Aizu, Japan, 2014. doi:10.1109/MCSoC.2014.31
    [BibTeX] [Abstract]

    Platforms are widely used to design embedded systems. They have numerous advantages: separation from its application, industrial rationalization, standardization, division of large development teams. However, their design complexity is growing dramatically due to several sources: the intricate combination of parallelism and heterogeneity in modern architectures; the quest for ever low power consumption and the diversity of sensors/actuators required by modern applications. This complexity prevents straightforward platform design in one step and calls for gradual design by composition and improvement over existing components. However, there is no systematic way of composing them, and there is no clear concept suitable for platform composition. In this paper, we propose two atomic ways of composing platforms, increment and assembly, that allow designers to build platforms gradually thanks to two concepts called world and container.

    @InProceedings{2014-09-HILI,
    author = {Nicolas Hili and Christian Fabre and Ivan Llopard and Sophie Dupuy-Chessa and Dominique Rieu},
    title = {{Model-Based Platform Composition for Embedded System Design}},
    booktitle = {{Proceedings of IEEE 8th International Symposium on Embedded Multicore Many-core Systems-on-Chip (MCSoC-14)}},
    date = {2014-09-23/2014-09-25},
    pages = {157-164},
    address = {University of Aizu, Japan},
    doi = {10.1109/MCSoC.2014.31},
    abstract = {Platforms are widely used to design embedded systems. They have numerous advantages: separation from its application, industrial rationalization, standardization, division of large development teams. However, their design complexity is growing dramatically due to several sources: the intricate combination of parallelism and heterogeneity in modern architectures; the quest for ever low power consumption and the diversity of sensors/actuators required by modern applications. This complexity prevents straightforward platform design in one step and calls for gradual design by composition and improvement over existing components. However, there is no systematic way of composing them, and there is no clear concept suitable for platform composition. In this paper, we propose two atomic ways of composing platforms, increment and assembly, that allow designers to build platforms gradually thanks to two concepts called world and container.},
    year = {2014}
    }

  • Nicolas Hili, Christian Fabre, Sophie Dupuy-Chessa, and Dominique Rieu. A Model-Driven Approach for Embedded System Prototyping and Design. In Proceedings of The IEEE International Symposium on Rapid System Prototyping (RSP), pages 23-29, New Delhi, India, 2014. doi:10.1109/RSP.2014.6966688
    [BibTeX] [Abstract]

    Embedded System (ES) development complexity is increasing. This increase has several cumulative sources: some are directly related to constraints on the ES themselves (dependability, compute intensive, resource constraints) while other sources are related to the industrial context of their development (fast prototyping, early validation, parallelization of developments). Although several Model-Driven Engineering (MDE) processes have been proposed for ES development, most of them are not completely formalized. This has several drawbacks that prevent their use in prototyping where iterations need to be short and focused. Incomplete formalized processes tend to be sidestepped in these situations where quick results are expected to be obtained with limited effort. In this paper we propose a MDE-based process for ES development. This process precisely defines the development tasks and their impact on the models throughout development. In particular we define iterations width and depth for the process that allow for a fined-grained and consistent planning of developments. The short and well defined iterations characterized by the process reduce the gap between rapid prototyping, ad-hoc methods and regular development processes.

    @InProceedings{2014-10-HILI,
    author = {Nicolas Hili and Christian Fabre and Sophie Dupuy-Chessa and Dominique Rieu},
    title = {{A Model-Driven Approach for Embedded System Prototyping and Design}},
    booktitle = {{Proceedings of The IEEE International Symposium on Rapid System Prototyping (RSP)}},
    date = {2014-10-16/2014-10-17},
    pages = {23-29},
    address = {New Delhi, India},
    doi = {10.1109/RSP.2014.6966688},
    abstract = {Embedded System (ES) development complexity is increasing. This increase has several cumulative sources: some are directly related to constraints on the ES themselves (dependability, compute intensive, resource constraints) while other sources are related to the industrial context of their development (fast prototyping, early validation, parallelization of developments). Although several Model-Driven Engineering (MDE) processes have been proposed for ES development, most of them are not completely formalized. This has several drawbacks that prevent their use in prototyping where iterations need to be short and focused. Incomplete formalized processes tend to be sidestepped in these situations where quick results are expected to be obtained with limited effort. In this paper we propose a MDE-based process for ES development. This process precisely defines the development tasks and their impact on the models throughout development. In particular we define iterations width and depth for the process that allow for a fined-grained and consistent planning of developments. The short and well defined iterations characterized by the process reduce the gap between rapid prototyping, ad-hoc methods and regular development processes.},
    year = {2014}
    }

  • Nicolas Hili. Une méthode pour le développement collaboratif de systèmes embarqués. PhD thesis, Université de Grenoble, 2014.
    [BibTeX] [Abstract] [Download PDF]

    Embedded system development is complex. This complexity has several sources. A first one is embedded system own specificities (physical world measurement and control, execution on a physical resource-constrained platform, reliability, responsiveness, …) that distinguish themselves from software systems. Another one comes from industrial concerns about whom these systems are subject to: product and development costs and delays, multidisciplinary teams, system documentation and certification. To handle this complexity, few methods and languages have been proposed. They focus on a modeling of both application and platform part included in an embedded system. However, the notions of method and process model are barely known from the embedded system community and current methods do not capitalize on the knowledge acquired by other engineering domains like information systems. The goal of this thesis is the definition, the formalization and the tooling of an embedded system development method. To do that, this work focuses on four main contributions: (1) the formalization of a guided process and a language to ensure a consistent modeling of both the application and the platform, (2) the composition of complex platforms to permit a progressive implementation of an application on its concrete platform, (3) the integration of a project management and a product traceability allowing the project manager to measure and monitor the development progress, to organize his team and to parallelize the development, and (4) the development of a tool designed to support the process, the language and the project management.

    @PhdThesis{2014-12-HILI,
    author = {Nicolas Hili},
    title = {{Une m\'{e}thode pour le d\'{e}veloppement collaboratif de syst\`{e}mes embarqu\'{e}s}},
    institution = {Universit\'{e} de Grenoble},
    date = {2014-12-11},
    url = {http://www.researchgate.net/publication/281377584_Une_mthode_pour_le_dveloppement_collaboratif_de_systmes_embarqus},
    address = {Universit\'{e} de Grenoble},
    abstract = {Embedded system development is complex. This complexity has several sources. A first one is embedded system own specificities (physical world measurement and control, execution on a physical resource-constrained platform, reliability, responsiveness, ...) that distinguish themselves from software systems. Another one comes from industrial concerns about whom these systems are subject to: product and development costs and delays, multidisciplinary teams, system documentation and certification. To handle this complexity, few methods and languages have been proposed. They focus on a modeling of both application and platform part included in an embedded system. However, the notions of method and process model are barely known from the embedded system community and current methods do not capitalize on the knowledge acquired by other engineering domains like information systems. The goal of this thesis is the definition, the formalization and the tooling of an embedded system development method. To do that, this work focuses on four main contributions: (1) the formalization of a guided process and a language to ensure a consistent modeling of both the application and the platform, (2) the composition of complex platforms to permit a progressive implementation of an application on its concrete platform, (3) the integration of a project management and a product traceability allowing the project manager to measure and monitor the development progress, to organize his team and to parallelize the development, and (4) the development of a tool designed to support the process, the language and the project management.},
    year = {2014}
    }

  • Andreas Erik Hindborg, Pascal Schleuniger, Nicklas Bo Jensen, and Sven Karlsson. Hardware Realization of an FPGA Processor – Operating System Call Offload and Experiences. In Proceedings of the Conference on Design and Architectures for Signal and Image Processing DASIP, Madrid, Spain, 2014. doi:10.1109/DASIP.2014.7115604
    [BibTeX] [Abstract]

    Field-programmable gate arrays, FPGAs, are attractive implementation platforms for low-volume signal and image processing applications. The structure of FPGAs allows for an efficient implementation of parallel algorithms. Sequential algorithms, on the other hand, often perform better on a microprocessor. It is therefore convenient for many applications to employ a synthesizable microprocessor to execute sequential tasks and custom hardware structures to accelerate parallel sections of an algorithm. In this paper, we discuss the hardware realization of Tinuso-I, a small synthesizable processor core that can be integrated in many signal and data processing platforms on FPGAs. We also show how we allow the processor to use operating system services. For a set of SPLASH-2 and SPEC CPU2006 benchmarks we show a speedup of up to 64% over a similar Xilinx MicroBlaze implementation while using 27% to 35% fewer hardware resources.

    @InProceedings{2014-10-HINDBORG,
    author = {Andreas Erik Hindborg and Pascal Schleuniger and Nicklas Bo Jensen and Sven Karlsson},
    title = {{Hardware Realization of an FPGA Processor - Operating System Call Offload and Experiences}},
    booktitle = {{Proceedings of the Conference on Design and Architectures for Signal and Image Processing {DASIP}}},
    date = {2014-10-8/2014-10-10},
    address = {Madrid, Spain},
    doi = {10.1109/DASIP.2014.7115604},
    abstract = {Field-programmable gate arrays, FPGAs, are attractive implementation platforms for low-volume signal and image processing applications. The structure of FPGAs allows for an efficient implementation of parallel algorithms. Sequential algorithms, on the other hand, often perform better on a microprocessor. It is therefore convenient for many applications to employ a synthesizable microprocessor to execute sequential tasks and custom hardware structures to accelerate parallel sections of an algorithm. In this paper, we discuss the hardware realization of Tinuso-I, a small synthesizable processor core that can be integrated in many signal and data processing platforms on FPGAs. We also show how we allow the processor to use operating system services. For a set of SPLASH-2 and SPEC CPU2006 benchmarks we show a speedup of up to 64% over a similar Xilinx MicroBlaze implementation while using 27% to 35% fewer hardware resources.},
    year = {2014}
    }

  • Andreas Erik Hindborg, Pascal Schleuniger, Nicklas Bo Jensen, Maxwell Walter, Laust Brock-Nannestad, Lars Bonnichsen, Christian W. Probst, and Sven Karlsson. Automatic Generation of Application Specific FPGA Multicore Accelerators. In 48th Asilomar conference on Signals Systems and Computers, Pacific Grove, USA, 2014. doi:10.1109/ACSSC.2014.7094700
    [BibTeX] [Abstract]

    High performance computing systems make increasing use of hardware accelerators to improve performance and power properties. For large high-performance FPGAs to be successfully integrated in such computing systems, methods to raise the abstraction level of FPGA programming are required. In this paper we propose a tool flow, which automatically generates highly optimized hardware multicore systems based on parameters. Profiling feedback is used to adjust these parameters to improve performance and lower the power consumption. For an image processing application we show that our tools are able to identify optimal performance energy trade-offs points for a multicore based FPGA accelerator

    @InProceedings{2014-11-HINDBORG,
    author = {Andreas Erik Hindborg and Pascal Schleuniger and Nicklas Bo Jensen and Maxwell Walter and Laust Brock-Nannestad and Lars Bonnichsen and Christian W. Probst and Sven Karlsson},
    title = {{Automatic Generation of Application Specific FPGA Multicore Accelerators}},
    booktitle = {{48th Asilomar conference on Signals Systems and Computers}},
    date = {2014-11-02/2014-11-05},
    address = {Pacific Grove, USA},
    doi = {10.1109/ACSSC.2014.7094700},
    abstract = {High performance computing systems make increasing use of hardware accelerators to improve performance and power properties. For large high-performance FPGAs to be successfully integrated in such computing systems, methods to raise the abstraction level of FPGA programming are required. In this paper we propose a tool flow, which automatically generates highly optimized hardware multicore systems based on parameters. Profiling feedback is used to adjust these parameters to improve performance and lower the power consumption. For an image processing application we show that our tools are able to identify optimal performance energy trade-offs points for a multicore based FPGA accelerator},
    year = {2014}
    }

  • Nicklas Bo Jensen, Christian W. Probst, and Sven Karlsson. Compiler Feedback using Continuous Dynamic Compilation during Development. In Proceedings of the Workshop on Dynamic Compilation Everywhere (DCE), Vienna, Austria, 2014.
    [BibTeX] [Abstract] [Download PDF]

    Optimizing compilers are vital for performance. However, compilers ability to optimize aggressively is limited in some cases. To address this limitation, we have developed a compiler guiding the programmer in making small source code changes, potentially making the source code more amenable to optimization. This tool can help programmers understand what the optimizing compiler has done and suggest automatic source code changes in cases where the compiler refrains from optimizing. We have integrated our tool into an integrated development environment, interactively giving feedback as part of the programmers development flow. We have evaluated our preliminary implementation and show it can guide to a 12% improvement in performance. Furthermore the tool can be used as an interactive optimization adviser improving the performance of the code generated by a production compiler. Here it can lead to a 153% improvement in performance, indicating the feasibility of the tool as a performance adviser for a production compiler.

    @InProceedings{2014-01-JENSEN,
    author = {Nicklas Bo Jensen and Christian W. Probst and Sven Karlsson},
    title = {{Compiler Feedback using Continuous Dynamic Compilation during Development}},
    booktitle = {{Proceedings of the Workshop on Dynamic Compilation Everywhere (DCE)}},
    address = {Vienna, Austria},
    date = {2014-01-21},
    url = {https://sites.google.com/site/dynamiccompilationeverywhere/home/dce-2014},
    abstract = {Optimizing compilers are vital for performance. However, compilers ability to optimize aggressively is limited in some cases. To address this limitation, we have developed a compiler guiding the programmer in making small source code changes, potentially making the source code more amenable to optimization.
    This tool can help programmers understand what the optimizing compiler has done and suggest automatic source code changes in cases where the compiler refrains from optimizing. We have integrated our tool into an integrated development environment, interactively giving feedback as part of the programmers development flow. We have evaluated our preliminary implementation and show it can guide to a 12% improvement in performance. Furthermore the tool can be used as an interactive optimization adviser improving the performance of the code generated by a production compiler. Here it can lead to a 153% improvement in performance, indicating the feasibility of the tool as a performance adviser for a production compiler.},
    year = {2014}
    }

  • Nicklas Bo Jensen, Sven Karlsson, Niklas Quarfot Nielsen, Gregory L. Lee, Dong H. Ahn, Matthew Legendre, and Martin Schulz. DySectAPI: Scalable Prescriptive Debugging. In Supercomputing, Portland, Oregon, USA, 2014.
    [BibTeX] [Abstract] [Download PDF]

    We present the DySectAPI, a tool that allow users to construct probe trees for automatic, event-driven debugging at scale. The traditional, interactive debugging model, whereby users manually step through and inspect their application, does not scale well even for current supercomputers. While lightweight debugging models scale well, they can currently only debug a subset of bug classes. DySectAPI fills the gap between these two approaches with a novel user-guided approach. Using both experimental results and analytical modeling we show how DySectAPI scales and can run with a low overhead on current systems.

    @InProceedings{2014-11-JENSEN-1,
    title = {{DySectAPI: Scalable Prescriptive Debugging}},
    author = {Nicklas Bo Jensen and Sven Karlsson and Niklas Quarfot Nielsen and Gregory L. Lee and Dong H. Ahn and Matthew Legendre and Martin Schulz},
    booktitle = {Supercomputing},
    address = {Portland, Oregon, USA},
    date = {2014-11-16/2014-11-21},
    url = {http://sc14.supercomputing.org/sites/all/themes/sc14/files/archive/tech_poster/tech_poster_pages/post237.html},
    year = 2014,
    abstract = {We present the DySectAPI, a tool that allow users to construct probe trees for automatic, event-driven debugging at scale. The traditional, interactive debugging model, whereby users manually step through and inspect their application, does not scale well even for current supercomputers. While lightweight debugging models scale well, they can currently only debug a subset of bug classes. DySectAPI fills the gap between these two approaches with a novel user-guided approach. Using both experimental results and analytical modeling we show how DySectAPI scales and can run with a low overhead on current systems.}
    }

  • Nicklas Bo Jensen, Christian W. Probst, and Sven Karlsson. Code Commentary and Automatic Refactorings using Feedback from Multiple Compilers. In Proceedings of the Swedish Workshop on Multicore Computing (MCC), Lund, Sweden, 2014.
    [BibTeX] [Abstract] [Download PDF]

    Optimizing compilers are essential to the performance of parallel programs on multi-core systems. It is attractive to expose parallelism to the compiler letting it do the heavy lifting. Unfortunately, it is hard to write code that compilers are able to optimize aggressively and therefore tools exist that can guide programmers with refactorings allowing the compilers to optimize more aggressively. We target the problem with many false positives that these tools often generate, where the amount of feedback can be overwhelming for the programmer. Our approach is to use a filtering scheme based on feedback from multiple compilers and show how we are able to filter out 87.6% of the comments by only showing the most promising comments.

    @InProceedings{2014-11-JENSEN-2,
    author = {Nicklas Bo Jensen and Christian W. Probst and Sven Karlsson},
    title = {{Code Commentary and Automatic Refactorings using Feedback from Multiple Compilers}},
    booktitle = {{Proceedings of the Swedish Workshop on Multicore Computing (MCC)}},
    date = {2014-11-27/2014-11-28},
    address = {Lund, Sweden},
    url = {http://orbit.dtu.dk/en/publications/code-commentary-and-automatic-refactorings-using-feedback-from-multiple-compilers(47725fbb-1c72-47fa-a167-32ae319d5a0d).html},
    abstract = {Optimizing compilers are essential to the performance of parallel programs on multi-core systems. It is attractive to expose parallelism to the compiler letting it do the heavy lifting. Unfortunately, it is hard to write code that compilers are able to optimize aggressively and therefore tools exist that can guide programmers with refactorings allowing the compilers to optimize more aggressively. We target the problem with many false positives that these tools often generate, where the amount of feedback can be overwhelming for the programmer. Our approach is to use a filtering scheme based on feedback from multiple compilers and show how we are able to filter out 87.6% of the comments by only showing the most promising comments.},
    year = {2014}
    }

  • Sandeep Katragadda, Juan C. SanMiguel, and Andrea Cavallaro. The costs of fusion in smart camera networks. In Proceedings of the International Conference on Distributed Smart Cameras, Venice, Italy, 2014. doi:10.1145/2659021.2659032
    [BibTeX] [Abstract]

    The choice of the most suitable fusion scheme for smart camera networks depends on the application as well as on the available computational and communication resources. In this paper we discuss and compare the resource requirements of five fusion schemes, namely centralised fusion, flooding, consensus, token passing and dynamic clustering. The Extended Information Filter is applied to each fusion scheme to perform target tracking. Token passing and dynamic clustering involve negotiation among viewing nodes (cameras observing the same target) to decide which node should perform the fusion process whereas flooding and consensus do not include this negotiation. Negotiation helps limiting the number of participating cameras and reduces the required resources for the fusion process itself but requires additional communication. Consensus has the highest communication and computation costs but it is the only scheme that can be applied when not all viewing nodes are connected directly and routing tables are not available.

    @InProceedings{2014-11-KATRAGADDA,
    title = {{The costs of fusion in smart camera networks}},
    author = {Sandeep Katragadda and Juan C. SanMiguel and Andrea Cavallaro},
    booktitle = {{Proceedings of the International Conference on Distributed Smart Cameras}},
    address= {Venice, Italy},
    date = {2014-11-04/2014-11-07},
    year = {2014},
    doi = {10.1145/2659021.2659032},
    abstract = {The choice of the most suitable fusion scheme for smart camera networks depends on the application as well as on the available computational and communication resources. In this paper we discuss and compare the resource requirements of five fusion schemes, namely centralised fusion, flooding, consensus, token passing and dynamic clustering. The Extended Information Filter is applied to each fusion scheme to perform target tracking. Token passing and dynamic clustering involve negotiation among viewing nodes (cameras observing the same target) to decide which node should perform the fusion process whereas flooding and consensus do not include this negotiation. Negotiation helps limiting the number of participating cameras and reduces the required resources for the fusion process itself but requires additional communication. Consensus has the highest communication and computation costs but it is the only scheme that can be applied when not all viewing nodes are connected directly and routing tables are not available.}
    }

  • Ivan Llopard, Albert Cohen, Christian Fabre, and Nicolas Hili. A Parallel Action Language for Embedded Applications and its Compilation Flow. In Proceedings of 17th International Workshop on Software and Compilers for Embedded Systems, pages 118-127, Schloss Rheinfels, St. Goar, Germany, 2014. doi:10.1145/2609248.2609257
    [BibTeX] [Abstract]

    The complexity of Embedded System (ES) development is increasing dramatically. This has several cumulative sources: the intricate combination of data-intensive, computational and control aspects; the ubiquity of parallelism and heterogeneity of modern architectures; and the diversity of target- specific, non-deterministic programming models (e.g., C++ with explicit message passing, OpenCL, VHDL). Model-Driven Engineering (MDE) proposes to manage complexity by raising the level of abstraction for designers and developers, and refining the implementation for a particular context and platform through model transformations. In such frameworks, behavior is often specified by means of Hierarchical State Machines (HSMs) equiped with an action language. However, although such models represent some level of control parallelism through objects and HSMs, data parallelism, compound data, and the exploitation and optimization thereof remains very limited. In this paper, we propose an action language that seamlessly combines HSMs with data parallelism and operations on compound data. It preserves the expressivity of HSM and captures a layout-neutral description of data organisation. It also extends message-passing with an intuitive semantics for this additional paralellism and provides strong foundation for array-based optimisation techniques. We present this language together with a baseline code generation flow to enable the production of efficient, low-level imperative code.

    @InProceedings{2014-06-LLOPARD,
    author = {Ivan Llopard and Albert Cohen and Christian Fabre and Nicolas Hili},
    title = {{A Parallel Action Language for Embedded Applications and its Compilation Flow}},
    booktitle = {{Proceedings of 17th International Workshop on Software and Compilers for Embedded Systems}},
    date = {2014-06-10/2014-06-11},
    address = {Schloss Rheinfels, St. Goar, Germany},
    pages = {118-127},
    doi = {10.1145/2609248.2609257},
    abstract = {The complexity of Embedded System (ES) development is increasing dramatically. This has several cumulative sources: the intricate combination of data-intensive, computational and control aspects; the ubiquity of parallelism and heterogeneity of modern architectures; and the diversity of target- specific, non-deterministic programming models (e.g., C++ with explicit message passing, OpenCL, VHDL). Model-Driven Engineering (MDE) proposes to manage complexity by raising the level of abstraction for designers and developers, and refining the implementation for a particular context and platform through model transformations. In such frameworks, behavior is often specified by means of Hierarchical State Machines (HSMs) equiped with an action language. However, although such models represent some level of control parallelism through objects and HSMs, data parallelism, compound data, and the exploitation and optimization thereof remains very limited. In this paper, we propose an action language that seamlessly combines HSMs with data parallelism and operations on compound data. It preserves the expressivity of HSM and captures a layout-neutral description of data organisation. It also extends message-passing with an intuitive semantics for this additional paralellism and provides strong foundation for array-based optimisation techniques. We present this language together with a baseline code generation flow to enable the production of efficient, low-level imperative code.},
    year = {2014}
    }

  • Prodromos-Vassilios Mekikis, Angelos Antonopoulos, and Christos Verikoukis. Wireless Energy Harvesting in Two-Way Network Coded Cooperative Communications: A Stochastic Approach for Large Scale Networks. IEEE Communication Letters, 18(6):1011-1014, 2014. doi:10.1109/LCOMM.2014.2320926
    [BibTeX] [Abstract]

    In this letter, we study the performance of network coding (NC)-aided cooperative communications in large scale networks, where the relays are able to harvest energy emitted by wireless transmissions. In particular, we derive theoretical expressions for key network performance metrics, i.e., the probability of successful data exchange and the network lifetime gain. The proposed analytical expressions are verified via extensive Monte Carlo simulations, demonstrating the potential benefits of the energy harvested by the wireless transmissions.

    @Article{2014-06-MEKIKIS-1,
    author = {Prodromos-Vassilios Mekikis and Angelos Antonopoulos and Christos Verikoukis},
    title = {{Wireless Energy Harvesting in Two-Way Network Coded Cooperative Communications: A Stochastic Approach for Large Scale Networks}},
    journal = {{IEEE Communication Letters}},
    date = {2014-06-06},
    volume = {18},
    number = {6},
    pages = {1011-1014},
    doi = {10.1109/LCOMM.2014.2320926},
    abstract = {In this letter, we study the performance of network coding (NC)-aided cooperative communications in large scale networks, where the relays are able to harvest energy emitted by wireless transmissions. In particular, we derive theoretical expressions for key network performance metrics, i.e., the probability of successful data exchange and the network lifetime gain. The proposed analytical expressions are verified via extensive Monte Carlo simulations, demonstrating the potential benefits of the energy harvested by the wireless transmissions.},
    year = {2014},
    }

  • Prodromos-Vassilios Mekikis, Elli Kartsakli, Angelos Antonopoulos, Aris S. Lalos, and Luis Alonso. Two-tier Cellular Random Network Planning for Minimum Deployment Cost. In Proceedings of the 2014 IEEE International Conference on Communications (ICC), pages 1248-1253, Sydney, Australia, 2014. doi:10.1109/ICC.2014.6883492
    [BibTeX] [Abstract]

    Random dense deployment of heterogeneous networks (HetNets), consisting of macro base stations (BS) and small cells (SC), can provide higher quality of service (QoS) while increasing the energy efficiency of the cellular network. In addition, it is possible to achieve lower deployment cost and, therefore, maximize the benefits for the network providers. In this paper, we propose a novel method to determine the minimum deployment cost of a two-tier heterogeneous cellular network using random deployment. After deriving the coverage probability of the two-tier deployment by using stochastic geometry tools, we identify the tier intensities that provide the minimum deployment cost for a given coverage probability. Extensive simulations verify the existence of a unique set of intensities for different coverage constraints.

    @InProceedings{2014-06-MEKIKIS-2,
    title = {{Two-tier Cellular Random Network Planning for Minimum Deployment Cost}},
    author = {Prodromos-Vassilios Mekikis and Elli Kartsakli and Angelos Antonopoulos and Aris S. Lalos and Luis Alonso},
    booktitle = {{Proceedings of the 2014 IEEE International Conference on Communications (ICC)}},
    date = {2014-06-10/2014-06-14},
    address = {Sydney, Australia},
    pages = {1248-1253},
    doi = {10.1109/ICC.2014.6883492},
    abstract = {Random dense deployment of heterogeneous networks (HetNets), consisting of macro base stations (BS) and small cells (SC), can provide higher quality of service (QoS) while increasing the energy efficiency of the cellular network. In addition, it is possible to achieve lower deployment cost and, therefore, maximize the benefits for the network providers. In this paper, we propose a novel method to determine the minimum deployment cost of a two-tier heterogeneous cellular network using random deployment. After deriving the coverage probability of the two-tier deployment by using stochastic geometry tools, we identify the tier intensities that provide the minimum deployment cost for a given coverage probability. Extensive simulations verify the existence of a unique set of intensities for different coverage constraints.},
    year = {2014}
    }

  • Vasileios Miliotis, Luis Alonso, and Christos Verikoukis. CooPNC: A cooperative multicast protocol exploiting physical layer network coding. Ad Hoc Networks, 14:35-50, 2014. doi:10.1016/j.adhoc.2013.11.004
    [BibTeX] [Abstract]

    In this paper we present a multicast protocol for short range networks that exploits the characteristics of physical layer network coding. In our proposed protocol, named CooPNC, we provide a novel cooperative approach that allows collision resolutions with the use of an indirect inter-network cooperation scheme. Through this scheme, we provide a reliable multicast protocol for partially overlapping short range networks with low control overhead. We show that with CooPNC higher throughput and energy efficiency are achieved, while it presents lower delay compared to state of the art multicast protocols. We provide an overview of our protocol with a simple scenario of overlapping networks and we then generalise its operation for a scalable scenario. Through mathematical analysis and simulations we prove that CooPNC presents significant performance gains compared to the state of the art in multicast communications for short range networks.

    @Article{2014-03-MILIOTIS,
    author = {Vasileios Miliotis and Luis Alonso and Christos Verikoukis},
    title = {{CooPNC: A cooperative multicast protocol exploiting physical layer network coding}},
    journal = {{Ad Hoc Networks}},
    volume = {14},
    pages = {35-50},
    date = {2014-03},
    doi = {10.1016/j.adhoc.2013.11.004},
    year = {2014},
    abstract = {In this paper we present a multicast protocol for short range networks that exploits the characteristics of physical layer network coding. In our proposed protocol, named CooPNC, we provide a novel cooperative approach that allows collision resolutions with the use of an indirect inter-network cooperation scheme. Through this scheme, we provide a reliable multicast protocol for partially overlapping short range networks with low control overhead. We show that with CooPNC higher throughput and energy efficiency are achieved, while it presents lower delay compared to state of the art multicast protocols. We provide an overview of our protocol with a simple scenario of overlapping networks and we then generalise its operation for a scalable scenario. Through mathematical analysis and simulations we prove that CooPNC presents significant performance gains compared to the state of the art in multicast communications for short range networks.}
    }

  • Raul Mohedano, Andrea Cavallaro, and Narciso Garcia. Camera localization using trajectories and maps. IEEE Transactions on Pattern Analysis and Machine Intelligence, 36(4):684-697, 2014. doi:10.1109/TPAMI.2013.243
    [BibTeX] [Abstract]

    We propose a new Bayesian framework for automatically determining the position (location and orientation) of an uncalibrated camera using the observations of moving objects and a schematic map of the passable areas of the environment. Our approach takes advantage of static and dynamic information on the scene structures through prior probability distributions for object dynamics. The proposed approach restricts plausible positions where the sensor can be located while taking into account the inherent ambiguity of the given setting. The proposed framework samples from the posterior probability distribution for the camera position via data driven MCMC, guided by an initial geometric analysis that restricts the search space. A Kullback-Leibler divergence analysis is then used that yields the final camera position estimate, while explicitly isolating ambiguous settings. The proposed approach is evaluated in synthetic and real environments, showing its satisfactory performance in both ambiguous and unambiguous settings.

    @Article{2014-04-MOHEDANO,
    title={{Camera localization using trajectories and maps}},
    author={Raul Mohedano and Andrea Cavallaro and Narciso Garcia},
    journal={{IEEE Transactions on Pattern Analysis and Machine Intelligence}},
    volume={36},
    number={4},
    pages={684-697},
    date={2014-04-01},
    year={2014},
    doi={10.1109/TPAMI.2013.243},
    abstract={We propose a new Bayesian framework for automatically determining the position (location and orientation) of an uncalibrated camera using the observations of moving objects and a schematic map of the passable areas of the environment. Our approach takes advantage of static and dynamic information on the scene structures through prior probability distributions for object dynamics. The proposed approach restricts plausible positions where the sensor can be located while taking into account the inherent ambiguity of the given setting. The proposed framework samples from the posterior probability distribution for the camera position via data driven MCMC, guided by an initial geometric analysis that restricts the search space. A Kullback-Leibler divergence analysis is then used that yields the final camera position estimate, while explicitly isolating ambiguous settings. The proposed approach is evaluated in synthetic and real environments, showing its satisfactory performance in both ambiguous and unambiguous settings.}
    }

  • Tahir Nawaz, Fabio Poiesi, and Andra Cavallaro. Assessing Tracking Assessment Measures. In Proceedings of the IEEE International Conference on Image Processing (ICIP), pages 441-445, Paris, France, 2014. doi:10.1109/ICIP.2014.7025088
    [BibTeX] [Abstract]

    We propose a methodology to quantitatively compare the relative performance of tracking evaluation measures. The proposed methodology is based on determining the probabilistic agreement between tracking result decisions made by measures and those made by humans. We use tracking results on publicly available datasets with different target types and varying challenges, and collect the judgments of 90 skilled, semi-skilled and unskilled human subjects using a web-based performance assessment test. The analysis of the agreements allows us to highlight the variation in performance of the different measures and the most appropriate ones for the various stages of tracking performance evaluation.

    @InProceedings{2014-10-NAWAZ,
    author = {Tahir Nawaz and Fabio Poiesi and Andra Cavallaro},
    title = {{Assessing Tracking Assessment Measures}},
    booktitle = {{Proceedings of the IEEE International Conference on Image Processing (ICIP)}},
    date = {2014-10-27/2014-10-30},
    address = {Paris, France},
    doi = {10.1109/ICIP.2014.7025088},
    pages = {441-445},
    abstract = {We propose a methodology to quantitatively compare the relative performance of tracking evaluation measures. The proposed methodology is based on determining the probabilistic agreement between tracking result decisions made by measures and those made by humans. We use tracking results on publicly available datasets with different target types and varying challenges, and collect the judgments of 90 skilled, semi-skilled and unskilled human subjects using a web-based performance assessment test. The analysis of the agreements allows us to highlight the variation in performance of the different measures and the most appropriate ones for the various stages of tracking performance evaluation.},
    year = {2014}
    }

  • Tahir Nawaz, Fabio Poiesi, and Andrea Cavallaro. Measures of effective video tracking. IEEE Transactions on Image Processing, 23(1):5-43, 2014. doi:10.1109/TIP.2013.2288578
    [BibTeX] [Abstract]

    To evaluate multitarget video tracking results, one needs to quantify the accuracy of the estimated target-size and the cardinality error as well as measure the frequency of occurrence of ID changes. In this paper, we survey existing multitarget tracking performance scores and, after discussing their limitations, we propose three parameter-independent measures for evaluating multitarget video tracking. The measures consider target-size variations, combine accuracy and cardinality errors, quantify long-term tracking accuracy at different accuracy levels, and evaluate ID changes relative to the duration of the track in which they occur. We conduct an extensive experimental validation of the proposed measures by comparing them with existing ones and by evaluating four state-of-the-art trackers on challenging real-world publicly-available data sets. The software implementing the proposed measures is made available online to facilitate their use by the research community.

    @Article{2014-01-NAWAZ-1,
    author = {Tahir Nawaz and Fabio Poiesi and Andrea Cavallaro},
    journal = {{IEEE Transactions on Image Processing}},
    title = {Measures of effective video tracking},
    date = {2014-01-01},
    volume = {23},
    pages = {5-43},
    number = {1},
    doi = {10.1109/TIP.2013.2288578},
    abstract = {To evaluate multitarget video tracking results, one needs to quantify the accuracy of the estimated target-size and the cardinality error as well as measure the frequency of occurrence of ID changes. In this paper, we survey existing multitarget tracking performance scores and, after discussing their limitations, we propose three parameter-independent measures for evaluating multitarget video tracking. The measures consider target-size variations, combine accuracy and cardinality errors, quantify long-term tracking accuracy at different accuracy levels, and evaluate ID changes relative to the duration of the track in which they occur. We conduct an extensive experimental validation of the proposed measures by comparing them with existing ones and by evaluating four state-of-the-art trackers on challenging real-world publicly-available data sets. The software implementing the proposed measures is made available online to facilitate their use by the research community.},
    year = {2014}
    }

  • Fabio Poiesi and Andrea Cavallaro. Predicting and recognizing Interactions in Public Spaces. Journal of Real-Time Image Processing, 10(4):785-803, 2014. doi:10.1007/s11554-014-0428-8
    [BibTeX] [Abstract]

    We present an extensive survey of methods for recognizing human interactions and propose a method for predicting rendezvous areas in observable and unobservable regions using sparse motion information. Rendezvous areas indicate where people are likely to interact with each other or with static objects (e.g., a door, an information desk or a meeting point). The proposed method infers the direction of movement by calculating prediction lines from displacement vectors and temporally accumulates intersecting locations generated by prediction lines. The intersections are then used as candidate rendezvous areas and modeled as spatial probability density functions using Gaussian Mixture Models. We validate the proposed method to predict dynamic and static rendezvous areas on real-world datasets and compare it with related approaches.

    @Article{2014-05-POIESI,
    author = {Fabio Poiesi and Andrea Cavallaro},
    title = {{Predicting and recognizing Interactions in Public Spaces}},
    journal = {{Journal of Real-Time Image Processing}},
    volume = {10},
    number = {4},
    pages = {785-803},
    date = {2014-05},
    doi = {10.1007/s11554-014-0428-8},
    abstract = {We present an extensive survey of methods for recognizing human interactions and propose a method for predicting rendezvous areas in observable and unobservable regions using sparse motion information. Rendezvous areas indicate where people are likely to interact with each other or with static objects (e.g., a door, an information desk or a meeting point). The proposed method infers the direction of movement by calculating prediction lines from displacement vectors and temporally accumulates intersecting locations generated by prediction lines. The intersections are then used as candidate rendezvous areas and modeled as spatial probability density functions using Gaussian Mixture Models. We validate the proposed method to predict dynamic and static rendezvous areas on real-world datasets and compare it with related approaches.},
    year = {2014}
    }

  • Fabio Poiesi and Andrea Cavallaro. Tracking Multiple High-Density Homogeneous Targets. IEEE Transactions on Circuits and Systems for Video Technology, 25(4):623-637, 2014. doi:10.1109/TCSVT.2014.2344509
    [BibTeX] [Abstract]

    We present a framework for multi-target detection and tracking that infers candidate target locations in videos containing a high density of homogeneous targets. We propose a gradient-climbing technique and an isocontour slicing approach for intensity maps to localize targets. The former uses Markov Chain Monte Carlo to iteratively fit a shape model onto the target locations, whereas the latter uses the intensity values at different levels to find consistent object shapes. We generate trajectories by recursively associating detections with a hierarchical graphbased tracker on temporal windows. The solution to the graph is obtained with a greedy algorithm that accounts for false positive associations. The edges of the graph are weighted with a likelihood function based on location information. We evaluate the performance of the proposed framework on challenging datasets containing videos with high density of targets and compare it with six alternative trackers.

    @Article{2014-07-POIESI-2,
    author = {Fabio Poiesi and Andrea Cavallaro},
    title = {{Tracking Multiple High-Density Homogeneous Targets}},
    journal = {{IEEE Transactions on Circuits and Systems for Video Technology}},
    date = {2014-07},
    volume = {25},
    number = {4},
    pages = {623-637},
    doi = {10.1109/TCSVT.2014.2344509},
    abstract = {We present a framework for multi-target detection and tracking that infers candidate target locations in videos containing a high density of homogeneous targets. We propose a gradient-climbing technique and an isocontour slicing approach for intensity maps to localize targets. The former uses Markov Chain Monte Carlo to iteratively fit a shape model onto the target locations, whereas the latter uses the intensity values at different levels to find consistent object shapes. We generate trajectories by recursively associating detections with a hierarchical graphbased tracker on temporal windows. The solution to the graph is obtained with a greedy algorithm that accounts for false positive associations. The edges of the graph are weighted with a likelihood function based on location information. We evaluate the performance of the proposed framework on challenging datasets containing videos with high density of targets and compare it with six alternative trackers.},
    year = {2014}
    }

  • Juan C. SanMiguel, Christian Micheloni, Karen Shoop, Gian Luca Foresti, and Andrea Cavallaro. Self-Reconfigurable Smart Camera Networks. IEEE Computer, 47(5):67-73, 2014. doi:10.1109/MC.2014.133
    [BibTeX] [Abstract]

    Camera networks that reconfigure while performing multiple tasks have unique requirements, such as concurrent task allocation with limited resources, the sharing of data among fields of view across the network, and coordination among heterogeneous devices.

    @Article{2014-05-SANMIGUEL,
    title={{Self-Reconfigurable Smart Camera Networks}},
    author={Juan C. SanMiguel and Christian Micheloni and Karen Shoop and Gian Luca Foresti and Andrea Cavallaro},
    journal={{IEEE Computer}},
    volume={47},
    number={5},
    pages={67-73},
    date={2014-05-20},
    doi={10.1109/MC.2014.133},
    year={2014},
    abstract={Camera networks that reconfigure while performing multiple tasks have unique requirements, such as concurrent task allocation with limited resources, the sharing of data among fields of view across the network, and coordination among heterogeneous devices.}
    }

  • Juan C. SanMiguel and Andrea Cavallaro. Cost-Aware Coalitions for Collaborative Tracking in Resource-Constrained Camera Networks. IEEE Sensors Journal, 15(5):2657-2668, 2014. doi:10.1109/JSEN.2014.2367015
    [BibTeX] [Abstract]

    We propose an approach to create camera coalitions in resource-constrained camera networks and demonstrate it for collaborative target tracking. We cast coalition formation as a decentralized resource allocation process where the best cameras among those viewing a target are assigned to a coalition based on marginal utility theory. A manager is dynamically selected to negotiate with cameras whether they will join the coalition and to coordinate the tracking task. This negotiation is based not only on the utility brought by each camera to the coalition, but also on the associated cost (i.e. additional processing and communication). Experimental results and comparisons using simulations and real data show that the proposed approach outperforms related state-of-the-art methods by improving tracking accuracy in cost-free settings. Moreover, under resource limitations, the proposed approach controls the tradeoff between accuracy and cost, and achieves energy savings with only a minor reduction in accuracy.

    @Article{2014-11-SANMIGUEL,
    title={{Cost-Aware Coalitions for Collaborative Tracking in Resource-Constrained Camera Networks}},
    author={Juan C. SanMiguel and Andrea Cavallaro},
    journal={{IEEE Sensors Journal}},
    volume={15},
    number={5},
    pages={2657-2668},
    date={2014-11-04},
    year={2014},
    doi={10.1109/JSEN.2014.2367015},
    abstract={We propose an approach to create camera coalitions in resource-constrained camera networks and demonstrate it for collaborative target tracking. We cast coalition formation as a decentralized resource allocation process where the best cameras among those viewing a target are assigned to a coalition based on marginal utility theory. A manager is dynamically selected to negotiate with cameras whether they will join the coalition and to coordinate the tracking task. This negotiation is based not only on the utility brought by each camera to the coalition, but also on the associated cost (i.e. additional processing and communication). Experimental results and comparisons using simulations and real data show that the proposed approach outperforms related state-of-the-art methods by improving tracking accuracy in cost-free settings. Moreover, under resource limitations, the proposed approach controls the tradeoff between accuracy and cost, and achieves energy savings with only a minor reduction in accuracy.}
    }

  • Grzegorz Szwoch. Parallel Background Subtraction in Video Streams using OpenCL on GPU Platforms. In 18th IEEE Conference SPA 2014: Signal Processing: Algorithms, Architectures, Arrangements, and Applications, pages 54-59, Poznan,Poland, 2014.
    [BibTeX] [Abstract] [Download PDF]

    Implementation of the background subtraction algorithm using OpenCL platform is presented. The algorithm processes live stream of video frames from the surveillance camera in on-line mode. Processing is performed using a host machine and a parallel computing device. The work focuses on optimizing an OpenCL algorithm implementation for GPU devices by taking into account specific features of the GPU architecture, such as memory access, data transfers and work group organization. However, the algorithm is intended to be used on any OpenCL compliant devices, including DSP and FPGA platforms. Various optimizations of the algorithm are presented and tested using a number of devices with varying processing power. The main aim of the work is to determine which optimizations are essential for ensuring on-line video processing in the surveillance system.

    @InProceedings{2014-09-SZWOCH,
    author = {Grzegorz Szwoch},
    title = {{Parallel Background Subtraction in Video Streams using OpenCL on GPU Platforms}},
    booktitle = {{18th IEEE Conference SPA 2014: Signal Processing: Algorithms, Architectures, Arrangements, and Applications}},
    date = {2014-09-22/2014-09-24},
    pages = {54-59},
    address = {Poznan,Poland},
    url = {http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber = 7067270&filter%3DAND(p_IS_Number%3A7067255)},
    abstract = {Implementation of the background subtraction algorithm using OpenCL platform is presented. The algorithm processes live stream of video frames from the surveillance camera in on-line mode. Processing is performed using a host machine and a parallel computing device. The work focuses on optimizing an OpenCL algorithm implementation for GPU devices by taking into account specific features of the GPU architecture, such as memory access, data transfers and work group organization. However, the algorithm is intended to be used on any OpenCL compliant devices, including DSP and FPGA platforms. Various optimizations of the algorithm are presented and tested using a number of devices with varying processing power. The main aim of the work is to determine which optimizations are essential for ensuring on-line video processing in the surveillance system.},
    year = {2014}
    }

  • Syed Fahad Tahir and Andrea Cavallaro. Cost-Effective Features for Reidentification in Camera Networks. IEEE Transactions on Circuits and Systems for Video Technology, 24(8):1362-1374, 2014. doi:10.1109/TCSVT.2014.2305511
    [BibTeX] [Abstract]

    Networks of smart cameras share large amounts of data to accomplish tasks such as reidentification. We propose a feature-selection method that minimizes the data needed to represent the appearance of objects by learning the most appropriate feature set for the task at hand (person reidentification). The computational cost for feature extraction and the cost for storing the feature descriptor are considered jointly with feature performance to select cost-effective good features. This selection allows us to improve intercamera reidentification while reducing the bandwidth that is necessary to share data across the camera network. We also rank the selected features in the order of effectiveness for the task to enable a further reduction of the feature set by dropping the least effective features when application constraints require this adaptation. We compare the proposed approach with state-of-the-art methods on the iLIDS and VIPeR datasets and show that the proposed approach considerably reduces network traffic due to intercamera feature sharing while keeping the reidentification performance at an equivalent or better level compared with the state of the art.

    @Article{2014-02-TAHIR,
    title={{Cost-Effective Features for Reidentification in Camera Networks}},
    author={Syed Fahad Tahir and Andrea Cavallaro},
    journal={{IEEE Transactions on Circuits and Systems for Video Technology}},
    volume={24},
    number={8},
    pages={1362-1374},
    date={2014-02-11},
    year={2014},
    doi={10.1109/TCSVT.2014.2305511},
    abstract={Networks of smart cameras share large amounts of data to accomplish tasks such as reidentification. We propose a feature-selection method that minimizes the data needed to represent the appearance of objects by learning the most appropriate feature set for the task at hand (person reidentification). The computational cost for feature extraction and the cost for storing the feature descriptor are considered jointly with feature performance to select cost-effective good features. This selection allows us to improve intercamera reidentification while reducing the bandwidth that is necessary to share data across the camera network. We also rank the selected features in the order of effectiveness for the task to enable a further reduction of the feature set by dropping the least effective features when application constraints require this adaptation. We compare the proposed approach with state-of-the-art methods on the iLIDS and VIPeR datasets and show that the proposed approach considerably reduces network traffic due to intercamera feature sharing while keeping the reidentification performance at an equivalent or better level compared with the state of the art.}
    }

  • Özge Ünel and Toygar Akgün. OpenCL Implementation of Unsharp Filtering on GPU and FPGA. In Proceedings of the 22nd IEEE Signal Processing and Communications Applications Conference (SIU 2014), pages 212-215, Karadeniz Technical University, Trabzon, Turkey, 2014. doi:10.1109/SIU.2014.6830203
    [BibTeX] [Abstract]

    The purpose of this study is to evaluate the performance of two dimensional multi-threaded linear filtering process on the GPU and FPGA platforms. To obtain the implementation on varying platforms, OpenCL API is used. OpenCL provides platform independent programming advantage. The results on three different platforms are compared to each other within this scope. These platforms are CPU, GPU, and FPGA. With changing filter and video frame sizes, varying processing times on these platforms are observed, and platform dependent advantages/disadvantages are studied.

    @InProceedings{2014-04-UNEL,
    author = {{\"O}zge {\"U}nel and Toygar Akg{\"u}n},
    title = {{OpenCL Implementation of Unsharp Filtering on GPU and FPGA}},
    booktitle = {{Proceedings of the 22nd IEEE Signal Processing and Communications Applications Conference (SIU 2014)}},
    date = {2014-04-23/2014-04-25},
    address = {Karadeniz Technical University, Trabzon, Turkey},
    pages = {212-215},
    doi = {10.1109/SIU.2014.6830203},
    abstract = {The purpose of this study is to evaluate the performance of two dimensional multi-threaded linear filtering process on the GPU and FPGA platforms. To obtain the implementation on varying platforms, OpenCL API is used. OpenCL provides platform independent programming advantage. The results on three different platforms are compared to each other within this scope. These platforms are CPU, GPU, and FPGA. With changing filter and video frame sizes, varying processing times on these platforms are observed, and platform dependent advantages/disadvantages are studied.},
    year = {2014}
    }

  • Maxwell Walter and Sven Karlsson. Testing Infrastructure for Operating System Kernel Development. In Proceedings of the Swedish Workshop on Multicore Computing (MCC), Lund, Sweden, 2014.
    [BibTeX] [Download PDF]
    @InProceedings{2014-11-WALTER,
    author = {Maxwell Walter and Sven Karlsson},
    title = {{Testing Infrastructure for Operating System Kernel Development}},
    booktitle = {{Proceedings of the Swedish Workshop on Multicore Computing (MCC)}},
    date = {2014-11-27/2014-11-28},
    address = {Lund, Sweden},
    url = {http://orbit.dtu.dk/fedora/objects/orbit:137313/datastreams/file_2596186f-b313-410f-8558-63b1c4a02261/content},
    year = {2014}
    }

  • Przemysaw Woznica, Michal Tarkowski, Marek Plotka, and Lukasz Kulas. RF Indoor Positioning System Supported by Wireless Computer Vision Sensors. In Proceedings of the 20th International Conference on Microwaves, Radar, and Wireless Communication (MIKON), pages 1-3, Gdansk, Poland, 2014.
    [BibTeX] [Abstract] [Download PDF]

    In this paper the possibility of increase the accuracy of RF fingerprinting indoor tracking system by the use of additional information from simple vision system is examined. As the distances in signal space differs from ones in real environment the ambiguity in decision process of fingerprinting algorithm can occur when set of closest distances between tag and map points in signal space corresponds to big distances differences in environment. The additional information from vision system is utilize to resolve which of possible locations should be chosen as correct one. In paper localization system is described and experimental results are presented as the evaluation of system performance.

    @InProceedings{2014-06-WOZNICA,
    author = {Przemysaw Woznica and Michal Tarkowski and Marek Plotka and Lukasz Kulas},
    title = {{RF Indoor Positioning System Supported by Wireless Computer Vision Sensors}},
    booktitle = {{Proceedings of the 20th International Conference on Microwaves, Radar, and Wireless Communication (MIKON)}},
    date = {2014-06-16/2014-06-18},
    pages = {1-3},
    address = {Gdansk, Poland},
    url = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber = 6899996&tag = 1},
    abstract = {In this paper the possibility of increase the accuracy of RF fingerprinting indoor tracking system by the use of additional information from simple vision system is examined. As the distances in signal space differs from ones in real environment the ambiguity in decision process of fingerprinting algorithm can occur when set of closest distances between tag and map points in signal space corresponds to big distances differences in environment. The additional information from vision system is utilize to resolve which of possible locations should be chosen as correct one. In paper localization system is described and experimental results are presented as the evaluation of system performance.},
    year = {2014}
    }

2013

  • Angelos Antonopoulos, Marco Di Renzo, and Christos Verikoukis. Effect of Realistic Channel Conditions on the Energy Efficiency of Network Coding-aided Cooperative MAC Protocols. Wireless Communications, IEEE, 20(5):76-84, 2013. doi:10.1109/MWC.2013.6664477
    [BibTeX] [Abstract]

    The emergence of heterogeneous networks (HetNets) as an enabling paradigm for ubiquitous wireless communication has further reinforced the concept of medium range cooperation among the end users. This trend, along with the need for bidirectional communication, has triggered the design of new Network Coding (NC)-aided Medium Access Control (MAC) protocols that benefit both the throughput and the energy efficiency in the system. However, the vast majority of MAC protocols are usually designed and analyzed under simplified channel models, ignoring the severe effect of realistic physical (PHY) layer conditions on the wireless communication. In this article, we focus on the impact of correlated long-term slow fading (shadowing) on the performance of distributed wireless systems. As a case study, we discuss in detail the performance of a cooperative NC-aided Automatic Repeat reQuest (ARQ) MAC protocol under correlated shadowing conditions. Our results reveal interesting trade-offs between throughput and energy efficiency, highlighting the importance of considering the slow fading effect in the design of cooperative MAC protocols

    @Article{2013-10-ANTONOPOULOS,
    author = {Angelos Antonopoulos and Marco Di Renzo and Christos Verikoukis},
    journal = {{Wireless Communications, IEEE}},
    title = {{Effect of Realistic Channel Conditions on the Energy Efficiency of Network Coding-aided Cooperative MAC Protocols}},
    date = {2013-10-01},
    volume = {20},
    number = {5},
    pages = {76-84},
    abstract = {The emergence of heterogeneous networks (HetNets) as an enabling paradigm for ubiquitous wireless communication has further reinforced the concept of medium range cooperation among the end users. This trend, along with the need for bidirectional communication, has triggered the design of new Network Coding (NC)-aided Medium Access Control (MAC) protocols that benefit both the throughput and the energy efficiency in the system. However, the vast majority of MAC protocols are usually designed and analyzed under simplified channel models, ignoring the severe effect of realistic physical (PHY) layer conditions on the wireless communication. In this article, we focus on the impact of correlated long-term slow fading (shadowing) on the performance of distributed wireless systems. As a case study, we discuss in detail the performance of a cooperative NC-aided Automatic Repeat reQuest (ARQ) MAC protocol under correlated shadowing conditions. Our results reveal interesting trade-offs between throughput and energy efficiency, highlighting the importance of considering the slow fading effect in the design of cooperative MAC protocols},
    year = {2013},
    doi = {10.1109/MWC.2013.6664477}
    }

  • Valentin Koblar and Bogdan Filipič. Designing a quality-control procedure for commutator manufacturing. In Proceedings of the 16th International Multiconference Information Society (IS 2013), volume A, pages 55-58, Ljubljana, Slovenia, 2013.
    [BibTeX] [Abstract] [Download PDF]

    This paper presents a methodology for designing an automated quality-control procedure in manufacturing graphite commutators for automotive industry. It focuses on assessing the quality of the copper-graphite joints on the commutators, using an advanced approach involving computer vision, machine learning and optimization. Based on the features of the captured images, classification models are built as a basis of the quality-control procedure, and optimized with respect to their accuracy and complexity. The preliminary results indicate that the proposed approach successfully identifies the features important for quality assessment, but needs further improvements to be applicable in industrial production.

    @InProceedings{2013-10-KOBLAR,
    author = {Valentin Koblar and Bogdan Filipi\v{c}},
    title = {Designing a quality-control procedure for commutator manufacturing},
    booktitle = {{Proceedings of the 16th International Multiconference Information Society (IS 2013)}},
    date = {2013-10-07/2013-10-11},
    volume = {A},
    pages = {55-58},
    address = {Ljubljana, Slovenia},
    url = {http://copcams.eu/wp-content/papercite-data/pdf/koblar13.pdf},
    abstract = {This paper presents a methodology for designing an automated quality-control procedure in manufacturing graphite commutators for automotive industry. It focuses on assessing the quality of the copper-graphite joints on the commutators, using an advanced approach involving computer vision, machine learning and optimization. Based on the features of the captured images, classification models are built as a basis of the quality-control procedure, and optimized with respect to their accuracy and complexity. The preliminary results indicate that the proposed approach successfully identifies the features important for quality assessment, but needs further improvements to be applicable in industrial production.},
    year = {2013}
    }

  • Thierry Lepley, Pierre Paulin, and Eric Flamand. A Novel Compilation Approach for Image Processing Graphs on a Many-Core Platform with Explicitly Managed Memory. In Embedded System Week 2013 (ESWeek) / International Conference on Compilers Architecture and Synthesis for Embedded Systems (CASES), Montréal, Canada, 2013. doi:10.1109/CASES.2013.6662510
    [BibTeX] [Abstract]

    Explicitly managed memory many-cores (EMM) have been a part of the industrial landscape for the last decade. The IBM CELL processor, general-purpose graphics processing units (GP-GPU) and the STHORM embedded many-core of STMicroelectronics are representative examples. This class of architecture is expected to scale well and to deliver good performance per watt and per mm2 of silicon. As such, it is appealing for application problems with regular data access patterns. However, this moves significant complexity to the programmer who must master parallelization and data movement. High level programming tools are therefore essential in order to allow the effective programming of EMM many-cores to a wide class of programmers. This paper presents a novel approach designed for simplifying the programming of EMM many-core architectures. It initially addresses the image processing application domain and has been targeted to the STHORM platform. It takes a high-level description of the computation kernel algorithm and generates an OpenCL kernel optimized for the target architecture, while managing the parallelization and data movements across the hierarchy in a transparent fashion. The goal is to provide both high productivity and high performance without requiring parallel computing expertise from the programmer, nor the need for application code specialization for the target architecture.

    @InProceedings{2013-09-LEPLEY,
    author = {Thierry Lepley and Pierre Paulin and Eric Flamand},
    title = {{A Novel Compilation Approach for Image Processing Graphs on a Many-Core Platform with Explicitly Managed Memory}},
    booktitle = {{Embedded System Week 2013 (ESWeek) / International Conference on Compilers Architecture and Synthesis for Embedded Systems (CASES)}},
    date = {2013-09-29/2013-10-04},
    address = {Montr\'{e}al, Canada},
    doi = {10.1109/CASES.2013.6662510},
    abstract = {Explicitly managed memory many-cores (EMM) have been a part of the industrial landscape for the last decade. The IBM CELL processor, general-purpose graphics processing units (GP-GPU) and the STHORM embedded many-core of STMicroelectronics are representative examples. This class of architecture is expected to scale well and to deliver good performance per watt and per mm2 of silicon. As such, it is appealing for application problems with regular data access patterns. However, this moves significant complexity to the programmer who must master parallelization and data movement. High level programming tools are therefore essential in order to allow the effective programming of EMM many-cores to a wide class of programmers. This paper presents a novel approach designed for simplifying the programming of EMM many-core architectures. It initially addresses the image processing application domain and has been targeted to the STHORM platform. It takes a high-level description of the computation kernel algorithm and generates an OpenCL kernel optimized for the target architecture, while managing the parallelization and data movements across the hierarchy in a transparent fashion. The goal is to provide both high productivity and high performance without requiring parallel computing expertise from the programmer, nor the need for application code specialization for the target architecture.},
    year = {2013}
    }

  • Ivan Llopard. Forces et faiblesses de LLVM comme générateur de code d’un DSP spécialisé. In Sixièmes rencontres de la communauté francaise de compilation, Annecy, France, 2013.
    [BibTeX] [Abstract] [Download PDF]

    Modern compilers integrate recent advances in compiler construction, intermediate representations, algorithms and programming language front-ends. Yet code generation for application-specific architectures benefits only marginally from this trend, as most of the effort is oriented towards popular general-purpose architectures. Historically, non-orthogonal architectures have relied on custom compiler technologies, some retargettable, but largely decoupled from the evolution of mainstream tool flows. Very Long Instruction Word (VLIW) architectures have introduced a variety of interesting problems such as clusterization, packetization or bundling, instruction scheduling for exposed pipelines, long delay slots, software pipelining, etc. These have been addressed in the literature, with a focus on the exploitation of Instruction Level Parallelism (ILP). While these are well known solutions already embedded into existing compilers, they rely on common hardware functionalities that are expected to be present in a fairly large subset of VLIW architectures. We’d like to present Mephisto, a high performance low-power application-specific processor, and discuss about the strenghts and weaknesses of LLVM as a code generator framework for this kind of processor. Mephisto is specialized enough to challenge established code generation solutions for VLIW and DSP processors, calling for an innovative compilation flow. Interesting features such as indexed register files, parallel address generation and exposed pipeline, among other specific functionalities, require dedicated compiler analysis and transformations in order to generate high-quality code.

    @InProceedings{2013-04-LLOPARD,
    author = {Ivan Llopard},
    title = {{Forces et faiblesses de LLVM comme g\'en\'erateur de code d'un DSP sp\'ecialis\'e}},
    booktitle = {{Sixi\`emes rencontres de la communaut\'e francaise de compilation}},
    date = {2013-04-02/2013-04-04},
    address = {Annecy, France},
    url = {http://compilfr.ens-lyon.fr/wp-content/uploads/2013/04/ivan.pdf},
    abstract = {Modern compilers integrate recent advances in compiler construction, intermediate representations, algorithms and programming language front-ends. Yet code generation for application-specific architectures benefits only marginally from this trend, as most of the effort is oriented towards popular general-purpose architectures. Historically, non-orthogonal architectures have relied on custom compiler technologies, some retargettable, but largely decoupled from the evolution of mainstream tool flows. Very Long Instruction Word (VLIW) architectures have introduced a variety of interesting problems such as clusterization, packetization or bundling, instruction scheduling for exposed pipelines, long delay slots, software pipelining, etc. These have been addressed in the literature, with a focus on the exploitation of Instruction Level Parallelism (ILP). While these are well known solutions already embedded into existing compilers, they rely on common hardware functionalities that are expected to be present in a fairly large subset of VLIW architectures. We'd like to present Mephisto, a high performance low-power application-specific processor, and discuss about the strenghts and weaknesses of LLVM as a code generator framework for this kind of processor. Mephisto is specialized enough to challenge established code generation solutions for VLIW and DSP processors, calling for an innovative compilation flow. Interesting features such as indexed register files, parallel address generation and exposed pipeline, among other specific functionalities, require dedicated compiler analysis and transformations in order to generate high-quality code. },
    year = {2013}
    }

  • Fabio Poiesi. Multi-target tracking and performance evaluation on videos. PhD thesis, Queen Mary University of London, 2013.
    [BibTeX] [Abstract] [Download PDF]

    Multi-target tracking is the process that allows the extraction of object motion patterns of interest from a scene. Motion patterns are often described through metadata representing object locations and shape information. In the first part of this thesis we discuss the state-of-the-art methods aimed at accomplishing this task on monocular views and also analyse the methods for evaluating their performance. The second part of the thesis describes our research contribution to these topics. We begin presenting a method for multi-target tracking based on track-before-detect (MTTBD) formulated as a particle filter. The novelty involves the inclusion of the target identity (ID) into the particle state, which enables the algorithm to deal with an unknown and unlimited number of targets. We propose a probabilistic model of particle birth and death based on Markov Random Fields. This model allows us to overcome the problem of the mixing of IDs of close targets. We then propose three evaluation measures that take into account target-size variations, combine accuracy and cardinality errors, quantify long-term tracking accuracy at different accuracy levels, and evaluate ID changes relative to the duration of the track in which they occur. This set of measures does not require pre-setting of parameters and allows one to holistically evaluate tracking performance in an application-independent manner. Lastly, we present a framework for multi-target localisation applied on scenes with a high density of compact objects. Candidate target locations are initially generated by extracting object features from intensity maps using an iterative method based on a gradient-climbing technique and an isocontour slicing approach. A graph-based data association method for multi-target tracking is then applied to link valid candidate target locations over time and to discard those which are spurious. This method can deal with point targets having indistinguishable appearance and unpredictable motion. MT-TBD is evaluated and compared with state-of-the-art methods on real-world surveillance

    @PhdThesis{2013-12-POIESI,
    author = {Fabio Poiesi},
    title = {{Multi-target tracking and performance evaluation on videos}},
    institution = {Universit\'{e} de Grenoble},
    date = {2013-12-11},
    url = {https://qmro.qmul.ac.uk/jspui/handle/123456789/8848},
    address = {Queen Mary University of London},
    abstract = {Multi-target tracking is the process that allows the extraction of object motion patterns of interest from a scene. Motion patterns are often described through metadata representing object locations and shape information. In the first part of this thesis we discuss the state-of-the-art methods aimed at accomplishing this task on monocular views and also analyse the methods for evaluating their performance. The second part of the thesis describes our research contribution to these topics. We begin presenting a method for multi-target tracking based on track-before-detect (MTTBD) formulated as a particle filter. The novelty involves the inclusion of the target identity (ID) into the particle state, which enables the algorithm to deal with an unknown and unlimited number of targets. We propose a probabilistic model of particle birth and death based on Markov Random Fields. This model allows us to overcome the problem of the mixing of IDs of close targets. We then propose three evaluation measures that take into account target-size variations, combine accuracy and cardinality errors, quantify long-term tracking accuracy at different accuracy levels, and evaluate ID changes relative to the duration of the track in which they occur. This set of measures does not require pre-setting of parameters and allows one to holistically evaluate tracking performance in an application-independent manner. Lastly, we present a framework for multi-target localisation applied on scenes with a high density of compact objects. Candidate target locations are initially generated by extracting object features from intensity maps using an iterative method based on a gradient-climbing technique and an isocontour slicing approach. A graph-based data association method for multi-target tracking is then applied to link valid candidate target locations over time and to discard those which are spurious. This method can deal with point targets having indistinguishable appearance and unpredictable motion. MT-TBD is evaluated and compared with state-of-the-art methods on real-world surveillance},
    year = {2013}
    }

Events

2016

  • Andrea Cavallaro. Autonomous robotic cameras for collaborative target localization. Invited talk at IEEE AVSS 2016 Workshop on Surveillance for Location-aware Data Protection, 2016.
    [BibTeX] [Download handouts]
    @Misc{2016-08-CAVALLARO,
    author = {Andrea Cavallaro},
    title = {{Autonomous robotic cameras for collaborative target localization}},
    howpublished = {Invited talk at IEEE AVSS 2016 Workshop on Surveillance for Location-aware Data Protection},
    date = {2016-08-23},
    year = {2016},
    address = {Colorado Springs, CO, USA},
    handouts = {http://www.eecs.qmul.ac.uk/~andrea/dwnld/2016.08.23_ColoradoSprings_AutonomousRoboticCameras.pdf}
    }

  • Andrea Cavallaro. Camera coalitions. Invited talk at IEEE-EURASIP Summer School on Signal Processing (S3P-2016), 2016.
    [BibTeX] [Abstract]

    Cameras are everywhere. Miniature high-quality cameras are increasingly worn by people, mounted on dashboards and micro-drones, omnipresent in hallways, streets and stores; and in your smartphone. Countless applications will benefit from the capabilities offered by networks of wireless cameras that can autonomously sense, compute, decide and communicate. These networks are composed of cameras whose algorithms need to adapt in response to unknown or dynamic environments and to changes in the assigned task. In this lecture I will present recent methods for cameras to move and to interact locally based on content and context, and to form coalitions that reach coordinated decisions under resource and physical constraints. I will discuss how cameras self-evaluate their performance and improve the quality of the task they are executing through collaboration, adaptively.

    @Misc{2016-09-CAVALLAROa,
    author = {Andrea Cavallaro},
    title = {{Camera coalitions}},
    howpublished = {Invited talk at IEEE-EURASIP Summer School on Signal Processing (S3P-2016)},
    date = {2016-09-04/2016-09-10},
    year = {2016},
    address = {Trento, IT},
    abstract = {Cameras are everywhere. Miniature high-quality cameras are increasingly worn by people, mounted on dashboards and micro-drones, omnipresent in hallways, streets and stores; and in your smartphone. Countless applications will benefit from the capabilities offered by networks of wireless cameras that can autonomously sense, compute, decide and communicate. These networks are composed of cameras whose algorithms need to adapt in response to unknown or dynamic environments and to changes in the assigned task. In this lecture I will present recent methods for cameras to move and to interact locally based on content and context, and to form coalitions that reach coordinated decisions under resource and physical constraints. I will discuss how cameras self-evaluate their performance and improve the quality of the task they are executing through collaboration, adaptively.}
    }

  • Andrea Cavallaro. Distributed Visual Processing. Tutorial at IEEE International Conference on Image Processing, 2016.
    [BibTeX] [Abstract]

    This tutorial will cover fundamental aspects, challenges and current solutions in distributed visual processing using networks of self-organising wired and wireless smart cameras, with applications in robotics, security and the Internet-of-Things. The tutorial sets forth the state-of-the-art in state estimation and coalition formation for distributed smart cameras. The tutorial will discuss and demonstrate the latest algorithms with a unified and comprehensive coverage. Using practical examples and illustration as support, the tutorial will introduce the participants in a discussion of the advantages and the limitations of traditional and modern approaches for synchronisation, distributed estimation and distributed processing for decision making and actuation in camera networks. Recent methods will be presented that allow cameras to move and to interact locally forming coalitions adaptively in order to provide coordinated decisions under resource and physical constraints. The tutorial will also discuss how cameras may learn to improve their performance. I will conclude the tutorial by introducing a collection of software resources to help the attendees develop and test distributed signal processing algorithms for wireless smart cameras.

    @Misc{2016-09-CAVALLAROb,
    author = {Andrea Cavallaro},
    title = {{Distributed Visual Processing}},
    howpublished = {Tutorial at IEEE International Conference on Image Processing},
    date = {2016-09-25/2016-09-28},
    year = {2016},
    address = {Phoenix, USA},
    abstract = {This tutorial will cover fundamental aspects, challenges and current solutions in distributed visual processing using networks of self-organising wired and wireless smart cameras, with applications in robotics, security and the Internet-of-Things. The tutorial sets forth the state-of-the-art in state estimation and coalition formation for distributed smart cameras. The tutorial will discuss and demonstrate the latest algorithms with a unified and comprehensive coverage. Using practical examples and illustration as support, the tutorial will introduce the participants in a discussion of the advantages and the limitations of traditional and modern approaches for synchronisation, distributed estimation and distributed processing for decision making and actuation in camera networks. Recent methods will be presented that allow cameras to move and to interact locally forming coalitions adaptively in order to provide coordinated decisions under resource and physical constraints. The tutorial will also discuss how cameras may learn to improve their performance. I will conclude the tutorial by introducing a collection of software resources to help the attendees develop and test distributed signal processing algorithms for wireless smart cameras.}
    }

  • Andrea Cavallaro. Smart Vision Meets Communication. Invited talk at Advanced Concepts for Intelligent Vision Systems 2016, 2016.
    [BibTeX]
    @Misc{2016-10-CAVALLAROa,
    author = {Andrea Cavallaro},
    title = {{Smart Vision Meets Communication}},
    howpublished = {Invited talk at Advanced Concepts for Intelligent Vision Systems 2016},
    date = {2016-10-24/2016-10-27},
    year = {2016},
    address = {Lecce, Italy}
    }

  • Andrea Cavallaro. Smart Cameras and Privacy. Invited talk at IC1206 Training School: De-identification for Privacy Protection in Multimedia Content, 2016.
    [BibTeX]
    @Misc{2016-10-CAVALLAROb,
    author = {Andrea Cavallaro},
    title = {{Smart Cameras and Privacy}},
    howpublished = {Invited talk at IC1206 Training School: De-identification for Privacy Protection in Multimedia Content},
    date = {2016-10-07/2016-10-11},
    year = {2016},
    address = {Limassol, Cyprus}
    }

  • François Duhem. Thales and the COPCAMS project: Efficient design of heterogeneous hardware and application optimization. Poster presented at the Embedded Vision Summit, 2016.
    [BibTeX]
    @Misc{2016-05-DUHEM,
    title = {{Thales and the COPCAMS project: Efficient design of heterogeneous hardware and application optimization}},
    author = {Fran\c{c}ois Duhem},
    date = {2016-05-02/2016-05-04},
    year = {2016},
    address = {Santa Clara, California, USA},
    howpublished = {Poster presented at the Embedded Vision Summit}
    }

  • Artur Podobas. Heterogeneous Computing with GPUs and FPGAs Through the OpenMP Task-Parallel Model. SIAM Conference on Parallel Processing for Scientific Computing, 2016.
    [BibTeX] [Abstract]

    The task-based programming model has become one the prominent models for parallelizing software on homogeneous shared-memory architectures. However, because Dennard’s scaling no longer holds, we have to leave homogeneity behind and move towards a more heterogeneous (possibly distributed) solution. In this talk, we will show how to extend the OpenMP task-model to support distributed computing through GPUs and how the model further can be used to drive hardware generation for FPGAs.

    @Misc{2016-04-PODOBAS,
    author = {Artur Podobas},
    title = {{Heterogeneous Computing with GPUs and FPGAs Through the OpenMP Task-Parallel Model}},
    howpublished = {SIAM Conference on Parallel Processing for Scientific Computing},
    date = {2016-04-12/2016-04-15},
    year = {2016},
    address = {Paris, France},
    organization = {Universit\'e Pierre \& Marie Curie},
    abstract = {The task-based programming model has become one the prominent models for parallelizing software on homogeneous shared-memory architectures. However, because Dennard's scaling no longer holds, we have to leave homogeneity behind and move towards a more heterogeneous (possibly distributed) solution. In this talk, we will show how to extend the OpenMP task-model to support distributed computing through GPUs and how the model further can be used to drive hardware generation for FPGAs.}
    }

  • Sven Verdoolaege, Tobias Grosser, and Albert Cohen. Polyhedral AST Generation is More than Scanning Polyhedra. Invited presentation at 37th annual ACM SIGPLAN conference on Programming Language Design and Implementation (PLDI), 2016.
    [BibTeX]
    @Misc{2016-06-VERDOOLAEGE,
    author = {Sven Verdoolaege and Tobias Grosser and Albert Cohen},
    title = {{Polyhedral {AST} Generation is More than Scanning Polyhedra}},
    date = {2016-06-15},
    howpublished = {Invited presentation at 37th annual ACM SIGPLAN conference on Programming Language Design and Implementation (PLDI)},
    address = {Santa Barbara, California, USA},
    date = {2016-06-13/2016-06-17},
    year = {2016}
    }

2015

  • Riyadh Baghdadi, Javed Absar, Adam Betts, Ulysse Beaugnon, Albert Cohen, Robert David, Alastair Donaldson, Tobias Grosser, Sven Van Haastregt, Elnar Hajiyev, Alexey Kravets, Jeroen Ketema, Michael Kruse, Anton Lokhmotov, Chandan Reddy, and Sven Verdoolaege. PENCIL: A Platform-Neutral Compute Intermediate Language for Parallelizing Compilation on Hardware Accelerators. HiPPES4CogApp (High-Performance Predictable Embedded Systems for Cognitive Applications workshop), colocated with HiPEAC 2015, 2015.
    [BibTeX]
    @Misc{2015-01-BAGHDADI,
    author = {Riyadh Baghdadi
    and Javed Absar
    and Adam Betts
    and Ulysse Beaugnon
    and Albert Cohen
    and Robert David
    and Alastair Donaldson
    and Tobias Grosser
    and Sven Van Haastregt
    and Elnar Hajiyev
    and Alexey Kravets
    and Jeroen Ketema
    and Michael Kruse
    and Anton Lokhmotov
    and Chandan Reddy
    and Sven Verdoolaege},
    title = {{PENCIL: A Platform-Neutral Compute Intermediate Language for Parallelizing Compilation on Hardware Accelerators}},
    howpublished = {HiPPES4CogApp (High-Performance Predictable Embedded Systems for Cognitive Applications workshop), colocated with HiPEAC 2015},
    date = {2015-01-22},
    address = {Amsterdam, The Netherlands},
    year = {2015}
    }

  • Andrea Cavallaro. Wearable and robotic camera networks. Keynote lecture at International Conference on Perception and Machine Intelligence, 2015.
    [BibTeX] [Abstract]

    The rapid progress in hardware, models and algorithms for smart camera networks will support the emergence of applications for the collaborative recognition of events using networked wearable cameras and networked camera-equipped unmanned vehicles (i.e. robotic camera networks). However, important challenges need to be addressed to enable wearable and robotic camera networks to impact on in-home assistive technologies, search and rescue missions, disaster management, routine monitoring and wildlife conservation. These challenges include efficient processing and decision making, adaptation to dynamic topology changes, coordination among heterogeneous devices, and the online evaluation of task performance. I this talk, I will discuss these challenges and present with several examples solutions from current and recent projects.

    @Misc{2015-02-CAVALLARO,
    author = {Andrea Cavallaro},
    title = {{Wearable and robotic camera networks}},
    howpublished = {Keynote lecture at International Conference on Perception and Machine Intelligence},
    date = {2015-02-26},
    address = {Kolkata, India},
    year = {2015},
    abstract = {The rapid progress in hardware, models and algorithms for smart camera networks will support
    the emergence of applications for the collaborative recognition of events using networked wearable
    cameras and networked camera-equipped unmanned vehicles (i.e. robotic camera networks). However,
    important challenges need to be addressed to enable wearable and robotic camera networks to impact on
    in-home assistive technologies, search and rescue missions, disaster management, routine monitoring and
    wildlife conservation. These challenges include efficient processing and decision making, adaptation to
    dynamic topology changes, coordination among heterogeneous devices, and the online evaluation of task
    performance. I this talk, I will discuss these challenges and present with several examples solutions from
    current and recent projects.}
    }

  • Andrea Cavallaro. Social Cameras. Keynote lecture at International Conference on Computer Vision, Imaging and Computer Graphics, 2015.
    [BibTeX] [Abstract]

    The pervasiveness of amateur wearable and handheld cameras has been revolutionizing the way private and public events are captured and shared. This explosion of content from social cameras and their shared user-generated videos offers the opportunity for new ways of analysing and timely reporting stories that range from music concerts and sports events to disaster scenes and protests. However, this large amount of increasingly available videos and their varying quality makes the selection and editing very difficult thus strongly limiting the opportunity to harvest the value of these user-generated videos. This talk will cover audio-visual methods for the automated grouping and synchronisation of multi-view user-generated videos in unconstrained scenarios. I will present ways to align these videos on a global timeline and show how to automatically edit them to produce a coherent final cut.

    @Misc{2015-03-CAVALLARO,
    author = {Andrea Cavallaro},
    title = {{Social Cameras}},
    howpublished = {Keynote lecture at International Conference on Computer Vision, Imaging and Computer Graphics},
    date = {2015-03-11},
    address = {Berlin, Germany},
    year = {2015},
    abstract = {The pervasiveness of amateur wearable and handheld cameras has been revolutionizing the way private and public events are captured and shared. This explosion of content from social cameras and their shared user-generated videos offers the opportunity for new ways of analysing and timely reporting stories that range from music concerts and sports events to disaster scenes and protests. However, this large amount of increasingly available videos and their varying quality makes the selection and editing very difficult thus strongly limiting the opportunity to harvest the value of these user-generated videos. This talk will cover audio-visual methods for the automated grouping and synchronisation of multi-view user-generated videos in unconstrained scenarios. I will present ways to align these videos on a global timeline and show how to automatically edit them to produce a coherent final cut.}
    }

  • Andrea Cavallaro. Privacy in the era of always-on everywhere cameras. Keynote lecture at Workshop on De-identification for privacy protection in multimedia in IEEE FG, 2015.
    [BibTeX] [Abstract]

    Video capturing devices are everywhere. Most of them are always on. Miniature high-quality cameras are increasingly worn by people, mounted on handlebars, dashboards and micro-drones, omnipresent in hallways, elevators, streets and stores; and in your smartphone. The pervasiveness of cameras in public places is offering incredible opportunities to improve our safety and quality of life, while at the same time posing important societal challenges. Enabling video-based applications and services while protecting privacy is a major interdisciplinary challenge that must be addressed to adapt to increased capabilities of technology and changing demands from citizens. In this talk I will cover multimedia applications that extract information from images, videos and camera networks as well as privacy constraints to be taken into account in the design of video-based systems. I will discuss basic principles and several application examples, and conclude with a set of open research questions.

    @Misc{2015-05-CAVALLARO,
    author = {Andrea Cavallaro},
    title = {{Privacy in the era of always-on everywhere cameras}},
    howpublished = {Keynote lecture at Workshop on De-identification for privacy protection in multimedia in IEEE FG},
    date = {2015-05-04},
    year = {2015},
    address = {Ljubljana, Slovenia},
    abstract = {Video capturing devices are everywhere. Most of them are always on. Miniature high-quality cameras are increasingly worn by people, mounted on handlebars, dashboards and micro-drones, omnipresent in hallways, elevators, streets and stores; and in your smartphone. The pervasiveness of cameras in public places is offering incredible opportunities to improve our safety and quality of life, while at the same time posing important societal challenges. Enabling video-based applications and services while protecting privacy is a major interdisciplinary challenge that must be addressed to adapt to increased capabilities of technology and changing demands from citizens. In this talk I will cover multimedia applications that extract information from images, videos and camera networks as well as privacy constraints to be taken into account in the design of video-based systems. I will discuss basic principles and several application examples, and conclude with a set of open research questions.}
    }

  • Andrea Cavallaro. Self-reconfiguring distributed vision. Plenary talk at International work conference on artificial neural networks (IWANN), 2015.
    [BibTeX] [Abstract]

    Assistive technologies, environmental monitoring, search and rescue operations, security and entertainment applications will considerably benefit from the sensing capabilities offered by emerging networks of wireless cameras. These networks are composed of cameras that may be wearable or mounted on robotic platforms and can autonomously sense, compute, decide and communicate. These cameras and their vision algorithms need to adapt their hardware and algorithmic parameters in response to unknown or dynamic environments and to changes in their task(s), i.e. they need to self-reconfigure. Cooperation among the cameras may lead to adaptive and task-dependent visual coverage of a scene or to increased robustness and accuracy in object localization under varying poses or illumination conditions. In this talk I will cover challenges and current solutions in self-reconfiguring distributed vision using networks of wireless cameras. In particular, I will discuss how cameras may learn to improve their performance. Moreover, I will present recent methods that allow cameras to move and to interact locally forming coalitions adaptively in order to provide coordinated decisions under resource and physical constraints.

    @Misc{2015-06-CAVALLARO,
    author = {Andrea Cavallaro},
    title = {{Self-reconfiguring distributed vision}},
    howpublished = {Plenary talk at International work conference on artificial neural networks (IWANN)},
    date = {2015-06-11},
    year = {2015},
    address = {Palma de Mallorca},
    abstract = {Assistive technologies, environmental monitoring, search and rescue operations, security and entertainment applications will considerably benefit from the sensing capabilities offered by emerging networks of wireless cameras. These networks are composed of cameras that may be wearable or mounted on robotic platforms and can autonomously sense, compute, decide and communicate. These cameras and their vision algorithms need to adapt their hardware and algorithmic parameters in response to unknown or dynamic environments and to changes in their task(s), i.e. they need to self-reconfigure. Cooperation among the cameras may lead to adaptive and task-dependent visual coverage of a scene or to increased robustness and accuracy in object localization under varying poses or illumination conditions. In this talk I will cover challenges and current solutions in self-reconfiguring distributed vision using networks of wireless cameras. In particular, I will discuss how cameras may learn to improve their performance. Moreover, I will present recent methods that allow cameras to move and to interact locally forming coalitions adaptively in order to provide coordinated decisions under resource and physical constraints.}
    }

  • Andrea Cavallaro. Smart Video Data. Keynote lecture at European Data Forum, 2015.
    [BibTeX] [Abstract]

    Video capturing devices are ubiquitous. Cameras with microphones are increasingly worn by people, mounted on dashboards and micro-drones; installed in hallways, streets and stores. In this talk I will cover techniques that automatically extract actionable information from videos and camera networks to learn patterns of behaviour, to detect anomalies and to help predict intentions. I will focus on recent applications that are offering extraordinary opportunities to improve our businesses, safety and quality of life; discuss the associated scientific challenges and privacy implications.

    @Misc{2015-11-CAVALLARO,
    author = {Andrea Cavallaro},
    title = {{Smart Video Data}},
    howpublished = {Keynote lecture at European Data Forum},
    date = {2015-11-16},
    year = {2015},
    address = {Luxembourg},
    abstract = {Video capturing devices are ubiquitous. Cameras with microphones are increasingly worn by people, mounted on dashboards and micro-drones; installed in hallways, streets and stores. In this talk I will cover techniques that automatically extract actionable information from videos and camera networks to learn patterns of behaviour, to detect anomalies and to help predict intentions. I will focus on recent applications that are offering extraordinary opportunities to improve our businesses, safety and quality of life; discuss the associated scientific challenges and privacy implications.}
    }

  • Henri-Pierre Charles. Binary Code Generation for Multimedia Application on Embedded Platforms. Advanced Concepts for Intelligent Vision Systems (ACIVS 2015), 2015.
    [BibTeX] [Abstract]

    Multimedia applications such as video compression, image processing, face recognition, run now on embedded platforms. The huge computing power needed is provided by the evolution of the transistor density and by using specialized accelerators. Theses accelerators are supported by multimedia instructions set. Using theses complex instructions can be a nightmare for the engineer because there is many way to program it, quality of the compiler support can be random depending on the couple compiler/platform and worse, performances can be data dependent. Using libraries can be an option if such library exist and provide enough performances. In this talk I’ll illustrate the difficulty to generate binary code for this application domain by practical example of code generation. Then I’ll show a tool deGoal which is developed in house to resolve these problems.

    @Misc{2015-10-CHARLES,
    author = {Henri-Pierre Charles},
    title = {{Binary Code Generation for Multimedia Application on Embedded Platforms}},
    howpublished = {Advanced Concepts for Intelligent Vision Systems (ACIVS 2015)},
    address = {Catania, Italy},
    date = {2015-10-26/2015-10-29},
    year = {2015},
    abstract = {Multimedia applications such as video compression, image processing, face recognition, run now on embedded platforms. The huge computing power needed is provided by the evolution of the transistor density and by using specialized accelerators. Theses accelerators are supported by multimedia instructions set.
    Using theses complex instructions can be a nightmare for the engineer because there is many way to program it, quality of the compiler support can be random depending on the couple compiler/platform and worse, performances can be data dependent. Using libraries can be an option if such library exist and provide enough performances.
    In this talk I'll illustrate the difficulty to generate binary code for this application domain by practical example of code generation. Then I'll show a tool deGoal which is developed in house to resolve these problems. }
    }

  • Luis Diaz and Eugenio Villar. VIPPE: Parallel Simulation and Performance Analysis of Complex Embedded Systems. HiPPES4CogApp (High-Performance Predictable Embedded Systems for Cognitive Applications workshop), colocated with HiPEAC 2015, 2015.
    [BibTeX]
    @Misc{2015-01-DIAZ,
    author = {Luis Diaz and Eugenio Villar},
    title = {{VIPPE: Parallel Simulation and Performance Analysis of Complex Embedded Systems}},
    howpublished = {HiPPES4CogApp (High-Performance Predictable Embedded Systems for Cognitive Applications workshop), colocated with HiPEAC 2015},
    date = {2015-01-20},
    address = {Amsterdam, Netherlands},
    year = {2015}
    }

  • Andreas Erik Hindborg, Nicklas Bo Jensen, Pascal Schleuniger, and Sven Karlsson. State of the Akvario Project. 6th Workshop on Architectural Research Prototyping (WARP 2015), 2015.
    [BibTeX] [Download PDF]
    @Misc{2015-06-HINDBORG,
    title = {{State of the Akvario Project}},
    author = {Andreas Erik Hindborg and Nicklas Bo Jensen and Pascal Schleuniger and Sven Karlsson},
    howpublished = {6th Workshop on Architectural Research Prototyping (WARP 2015)},
    address = {Portland, Oregon, USA},
    date = {2015-06-14},
    year = {2015},
    url = {http://www.csl.cornell.edu/warp2015/abstracts/hindborg-akvario-warp2015.pdf}
    }

  • Andreas Erik Hindborg, Lars Frydendal Bonnichsen, Nicklas Bo Jensen, Laust Brock-Nannestad, Christian W. Probst, and Sven Karlsson. An Adaptive Framework for Imaging Systems. Advanced Concepts for Intelligent Vision Systems (ACIVS 2015), 2015.
    [BibTeX] [Abstract]

    Computer vision and video processing systems handle large amounts of data with varying spatial and temporal resolution and multiple imaging modalities. The current best practice is to design video processing systems with an overcapacity, which avoids underperforming in the general case, but wastes resources. In this work we present an adaptive framework for imaging systems that aims at minimizing waste of resources. Depending on properties of the processed images, the system dynamically adapts both the implementation of the processing system and properties of the underlying hardware.

    @Misc{2015-10-HINDBORG,
    author = {Andreas Erik Hindborg and Lars Frydendal Bonnichsen and Nicklas Bo Jensen and Laust Brock-Nannestad and Christian W. Probst and Sven Karlsson},
    title = {{An Adaptive Framework for Imaging Systems}},
    howpublished = {Advanced Concepts for Intelligent Vision Systems (ACIVS 2015)},
    address = {Catania, Italy},
    date = {2015-10-26/2015-10-29},
    year = {2015},
    abstract = {Computer vision and video processing systems handle large amounts of data with varying spatial and temporal resolution and multiple imaging modalities. The current best practice is to design video processing systems with an overcapacity, which avoids underperforming in the general case, but wastes resources. In this work we present an adaptive framework for imaging systems that aims at minimizing waste of resources. Depending on properties of the processed images, the system dynamically adapts both the implementation of the processing system and properties of the underlying hardware. }
    }

  • Maxwell Walter, Pascal Schleuniger, Andreas Erik Hindborg, Carl Christian Kjaergaard, Nicklas Bo Jensen, and Sven Karlsson. Experiences Implementing Tinuso in gem5. Second gem5 User Workshop, 2015.
    [BibTeX] [Abstract] [Download handouts]

    In recent years, high performance computing systems have started to make use of FPGA based hardware accelerators to improve performance and power properties. While FPGAs are becoming more competitive in terms of speed, power efficiency, and logic capacity, the effort required to program these heterogeneous systems has limited their impact. To address these issues, our lab has developed Tinuso, which is is a lightweight processor architecture we designed and optimized for implementation on FPGAs. Tinuso is designed to be multi-core and configurable for specific applications. To identify an ideal multi-core configuration for an application, we need a powerful simulation environment that can efficiently explore the design space. We use gem5 as our simulation platform, and have added support for our processor architecture. For Tinuso, we follow a hardware/software co-design approach in order to keep the hardware resource usage low. This results in processor cores that deliver significantly higher performance while requiring fewer hardware resources than commer- cial processor implementations. However, Tinuso is in many ways different from other processor architectures. To maintain high operating frequencies Tinuso includes a large number of delay slots; four for standard branch instructions. Tinuso also includes delay slots on other instructions, where the result is not available for a number of cycles after the instruction has been executed. To complicate matters, certain instruction patterns have a different number of delay slots. For example, the compare instruction normally has two delay slots, but only one when followed by a branch. We have modified the gem5 in-order CPU model to support a dynamic commit delay. This allows us to fetch instructions in the proper order, which maintains cache behavior, but still simulate the behavior of Tinuso?s delay slots. Along the way, we have found gem5 to be a very flexible platform that we have been able to use for our design space exploration and compiler verification.

    @Misc{2015-06-MAXWELL,
    title = {{Experiences Implementing Tinuso in gem5}},
    author = {Maxwell Walter and Pascal Schleuniger and Andreas Erik Hindborg and Carl Christian Kjaergaard and Nicklas Bo Jensen and Sven Karlsson},
    howpublished = {Second gem5 User Workshop},
    address = {Portland, Oregon, USA},
    date = {2015-06-14},
    year = {2015},
    handouts = {http://www.m5sim.org/wiki/images/f/f5/2015_ws_16_gem5-workshop_mwalter.pptx},
    abstract = {In recent years, high performance computing systems have started to make use of FPGA based hardware accelerators to improve performance and power properties. While FPGAs are becoming more competitive in terms of speed, power efficiency, and logic capacity, the effort required to program these heterogeneous systems has limited their impact. To address these issues, our lab has developed Tinuso, which is is a lightweight processor architecture we designed and optimized for implementation on FPGAs. Tinuso is designed to be multi-core and configurable for specific applications. To identify an ideal multi-core configuration for an application, we need a powerful simulation environment that can efficiently explore the design space. We use gem5 as our simulation platform, and have added support for our processor architecture.
    For Tinuso, we follow a hardware/software co-design approach in order to keep the hardware resource usage low. This results in processor cores that deliver significantly higher performance while requiring fewer hardware resources than commer- cial processor implementations. However, Tinuso is in many ways different from other processor architectures. To maintain high operating frequencies Tinuso includes a large number of delay slots; four for standard branch instructions.
    Tinuso also includes delay slots on other instructions, where the result is not available for a number of cycles after the instruction has been executed. To complicate matters, certain instruction patterns have a different number of delay slots. For example, the compare instruction normally has two delay slots, but only one when followed by a branch. We have modified the gem5 in-order CPU model to support a dynamic commit delay. This allows us to fetch instructions in the proper order, which maintains cache behavior, but still simulate the behavior of Tinuso?s delay slots. Along the way, we have found gem5 to be a very flexible platform that we have been able to use for our design space exploration and compiler verification.}
    }

  • Anca Molnos. System-Level Energy Management in Many-Core Systems Utilising Distributed Speed-Power Controllers. Invited Lecture at COOLCHIPS 2015 (XVIIIth IEEE Symposium on Low-Power and High-Speed Chips), 2015.
    [BibTeX] [Abstract]

    Energy efficiency is one of the crucial concerns today in computing systems ranging from small connected devices to large data-centers. This issue is addressed a various levels, and recently we have witnessed a lot of progress in methods to control speed and power consumption of digital circuits. Notable examples are fine-grain adaptive voltage and frequency scaling, and the adoption of new technologies such as Fully-Depleted Silicon On Insulator (FDSOI). These advances however bring new knobs to tradeoff power and speed, e.g., supply voltage, body-bias voltage, which, in turn, open interesting questions about how to fully take advantage of their potential at software level. This talk we will present methods to reduce power consumption of applications and the tradeoffs therein. As a research vehicle, we have the case of a low-power many-core architecture with several power domains and distributed speed-power controllers. We will study the impact of adaptive voltage scaling and discuss methods to determine the optimal power modes, both with benefits at system level, in the context of advanced technologies such as FDSOI.

    @Misc{2015-04-MOLNOS,
    author = {Anca Molnos},
    title = {{System-Level Energy Management in Many-Core Systems Utilising Distributed Speed-Power Controllers}},
    howpublished = {Invited Lecture at COOLCHIPS 2015 (XVIIIth IEEE Symposium on Low-Power and High-Speed Chips)},
    date = {2015-04-13},
    address = {Yokohama, Japan},
    abstract = {Energy efficiency is one of the crucial concerns today in computing systems ranging from small connected devices to large data-centers. This issue is addressed a various levels, and recently we have witnessed a lot of progress in methods to control speed and power consumption of digital circuits. Notable examples are fine-grain adaptive voltage and frequency scaling, and the adoption of new technologies such as Fully-Depleted Silicon On Insulator (FDSOI). These advances however bring new knobs to tradeoff power and speed, e.g., supply voltage, body-bias voltage, which, in turn, open interesting questions about how to fully take advantage of their potential at software level. This talk we will present methods to reduce power consumption of applications and the tradeoffs therein. As a research vehicle, we have the case of a low-power many-core architecture with several power domains and distributed speed-power controllers. We will study the impact of adaptive voltage scaling and discuss methods to determine the optimal power modes, both with benefits at system level, in the context of advanced technologies such as FDSOI.},
    year = {2015}
    }

  • Anca Molnos. SoC Energy Management via Feedback Control for Data Parallel Applications with Throughput Constraints. MPSoC 2015, 2015.
    [BibTeX]
    @Misc{2015-07-MOLNOS,
    author = {Anca Molnos},
    title = {{SoC Energy Management via Feedback Control for Data Parallel Applications with Throughput Constraints}},
    howpublished = {MPSoC 2015},
    address = {Ventura Beach, California, USA},
    date = {2015-07-13/2015-07-17},
    year = {2015}
    }

  • Andrew O’Byrne, Charles Attwood, and Darryl Hond. Porting and optimisation of the Thales robust motion detector to the iMX6 platform. HiPPES4CogApp (High-Performance Predictable Embedded Systems for Cognitive Applications workshop), colocated with HiPEAC 2015, 2015.
    [BibTeX]
    @Misc{2015-01-OBYRNE,
    author = {Andrew O'Byrne and Charles Attwood and Darryl Hond},
    title = {{Porting and optimisation of the Thales robust motion detector to the iMX6 platform}},
    howpublished = {HiPPES4CogApp (High-Performance Predictable Embedded Systems for Cognitive Applications workshop), colocated with HiPEAC 2015},
    date = {2015-01-20},
    address = {Amsterdam, The Netherlands},
    year = {2015}
    }

  • Michael Tchagaspanian. Smart Image Sensor for Advanced Use and New Applications. Advanced Concepts for Intelligent Vision Systems (ACIVS 2015), 2015.
    [BibTeX] [Abstract]

    Multimedia applications such as video compression, image processing, face recognition, run now on embedded platforms. The huge computing power needed is provided by the evolution of the transistor density and by using specialized accelerators. These accelerators are supported by multimedia instructions set. Using these complex instructions can be a nightmare for the engineer because there are many ways to program it, quality of the compiler support can be random depending on the couple compiler/platform and worse, performances can be data dependent. Using libraries can be an option if such libraries exist and provide enough performances. In this talk, I’ll illustrate the difficulty to generate binary code for this application domain by practical example of code generation. Then I’ll show a tool deGoal which is developed in house to resolve these problems.

    @Misc{2015-10-TCHAGASPANIAN,
    author = {Michael Tchagaspanian},
    title = {{Smart Image Sensor for Advanced Use and New Applications}},
    howpublished = {Advanced Concepts for Intelligent Vision Systems (ACIVS 2015)},
    address = {Catania, Italy},
    date = {2015-10-26/2015-10-29},
    year = {2015},
    abstract = {Multimedia applications such as video compression, image processing, face recognition, run now on embedded platforms. The huge computing power needed is provided by the evolution of the transistor density and by using specialized accelerators. These accelerators are supported by multimedia instructions set.
    Using these complex instructions can be a nightmare for the engineer because there are many ways to program it, quality of the compiler support can be random depending on the couple compiler/platform and worse, performances can be data dependent. Using libraries can be an option if such libraries exist and provide enough performances.
    In this talk, I'll illustrate the difficulty to generate binary code for this application domain by practical example of code generation. Then I'll show a tool deGoal which is developed in house to resolve these problems. }
    }

  • Özge Ünel and Toygar Akgün. OpenCL Implementation of Mean Shift Algorithm with Performance Comparisons on Discrete and Embedded . Advanced Concepts for Intelligent Vision Systems (ACIVS 2015), 2015.
    [BibTeX]
    @Misc{2015-10-UNEL,
    author = {\"{O}zge \"{U}nel and Toygar Akg\"{u}n},
    title = {{OpenCL Implementation of Mean Shift Algorithm with Performance Comparisons on Discrete and Embedded }},
    howpublished = {Advanced Concepts for Intelligent Vision Systems (ACIVS 2015)},
    address = {Catania, Italy},
    date = {2015-10-26/2015-10-29},
    year = {2015}
    }

2014

  • Ulysse Beaugnon, Riyadh Baghdadi, Javed Absar, Adam Betts, Albert Cohen, Alastair Donaldson, Tobias Grosser, Sven van Haastregt, Yabin Hu, Jeroen Ketema, Alexey Kravets, Anton Lokhmotov, and Sven Verdoolaege. PENCIL: A platform-neutral intermediate language for the parallelizing compilation of DSLs. Second Workshop on Domain-Specific Language Design and Implementation (DSLDI 2014), 2014. Held in conjunction with SPLASH/OOPSLA 2014
    [BibTeX]
    @Misc{2014-10-BEAUGNON,
    author = {Ulysse Beaugnon and Riyadh Baghdadi and Javed Absar
    and Adam Betts and Albert Cohen and Alastair Donaldson
    and Tobias Grosser and Sven van Haastregt and Yabin Hu
    and Jeroen Ketema and Alexey Kravets and
    Anton Lokhmotov and Sven Verdoolaege},
    title = {{PENCIL: A platform-neutral intermediate language for the parallelizing compilation of DSLs}},
    howpublished = {Second Workshop on Domain-Specific Language Design and Implementation (DSLDI 2014)},
    note = {Held in conjunction with SPLASH/OOPSLA 2014},
    date = {2014-10-19/2014-10-24},
    address = {Portland, OR, USA},
    year = {2014}
    }

  • Ayoub Nouri, Anca Molnos, Marius Bozga, Julien Mottin, Arnaud Tonda, Saddek Bensalem, and François Pacull. A Model-based Approach for Rapid Prototyping of Parallel Applications on Many-core. Poster presented at the 51st Design Automation Conference in San Francisco, USA, 2014.
    [BibTeX] [Abstract]

    Rapid prototyping of highly parallel applications on many-core platforms is extremely challenging. We present an automated analysis and code generation flow for implementing high-level KPN models on STHORM, an embedded 64-core computing fabric developed by STMicroelectronics. The flow is model-based with sound semantical basis and enables formal verification and performance analysis at different stages. The target for code generation is the MCAPI programming standard which has been ported on STHORM. Currently, the flow is fully operational. We report concrete results obtained on image-processing algorithms and illustrate the potential benefits of the flow for exploring implementation trade-offs.

    @Misc{2014-06-NOURI,
    author = {Ayoub Nouri and Anca Molnos and Marius Bozga and Julien Mottin and Arnaud Tonda and Saddek Bensalem and Fran\c{c}ois Pacull},
    title = {{A Model-based Approach for Rapid Prototyping of Parallel Applications on Many-core}},
    date = {2014-06-01/2014-06-05},
    howpublished = {Poster presented at the 51st Design Automation Conference in San Francisco, USA},
    abstract = {Rapid prototyping of highly parallel applications on many-core platforms is extremely challenging. We present an automated analysis and code generation flow for implementing high-level KPN models on STHORM, an embedded 64-core computing fabric developed by STMicroelectronics. The flow is model-based with sound semantical basis and enables formal verification and performance analysis at different stages. The target for code generation is the MCAPI programming standard which has been ported on STHORM. Currently, the flow is fully operational. We report concrete results obtained on image-processing algorithms and illustrate the potential benefits of the flow for exploring implementation trade-offs.},
    year = {2014}
    }

  • Tiana Rakotovao, Maxime Louvel, Anca Molnos, Julien Mottin, and François Pacull. Distributed Coordination of Sub-systems Power-modes and Software-modes. Poster presented at the 5th International Conference on Cyber-Physical Systems (ICCPS) in Berlin, Germany, 2014.
    [BibTeX] [Abstract]

    Energy management is essential for cyber-physical systems. Such systems typically consist of several sub-systems that may communicate. Hardware used in these sub-systems often has several power-modes that can be controlled to consume less energy. However, to-date, the decision of power-mode does not consider neither the external context of the system, nor the software-modes which involve on the QoS of the system. To address this problem, we propose a loosely coupled and distributed framework that selects the appropriate sub-systems power-modes, regarding both external context (e.g. GPS location, ambient temperature, information from external applications) and software-modes. The framework is based on the LINC coordination middleware. It is evaluated in a vehicle obstacle perception application running on the STHORM platform, a manycore SoC. Depending on the vehicle location and speed, the application has different software-mode and processing requirements. This information is used to scale the power-mode of STHORM in order to reduce energy consumption. The results have shown that significant power-saving can be realized by considering external context. The loosely coupled approach of LINC simplifies the integration of any sub-systems into the framework. In addition, through its goal-driven production rules, LINC allows to easily coordinate the whole system and to adapt it to the external context.

    @Misc{2014-04-RAKOTOVAO,
    author = {Tiana Rakotovao and Maxime Louvel and Anca Molnos and Julien Mottin and Fran\c{c}ois Pacull},
    title = {{Distributed Coordination of Sub-systems Power-modes and Software-modes}},
    date = {2014-04-14/2014-04-17},
    howpublished = {Poster presented at the 5th International Conference on Cyber-Physical Systems (ICCPS) in Berlin, Germany},
    abstract = {Energy management is essential for cyber-physical systems. Such systems typically consist of several sub-systems that may communicate. Hardware used in these sub-systems often has several power-modes that can be controlled to consume less energy. However, to-date, the decision of power-mode does not consider neither the external context of the system, nor the software-modes which involve on the QoS of the system. To address this problem, we propose a loosely coupled and distributed framework that selects the appropriate sub-systems power-modes, regarding both external context (e.g. GPS location, ambient temperature, information from external applications) and software-modes. The framework is based on the LINC coordination middleware. It is evaluated in a vehicle obstacle perception application running on the STHORM platform, a manycore SoC. Depending on the vehicle location and speed, the application has different software-mode and processing requirements. This information is used to scale the power-mode of STHORM in order to reduce energy consumption. The results have shown that significant power-saving can be realized by considering external context. The loosely coupled approach of LINC simplifies the integration of any sub-systems into the framework. In addition, through its goal-driven production rules, LINC allows to easily coordinate the whole system and to adapt it to the external context.},
    year = {2014}
    }

  • Maxwell Walter and Sven Karlsson. Unit Testing Framework for Operating System Kernels. Poster at the USENIX Symposium on Operating Systems Design and Implementation (OSDI), 2014.
    [BibTeX] [Download PDF]
    @Misc{2014-10-WALTER,
    author = {Maxwell Walter and Sven Karlsson},
    title = {{Unit Testing Framework for Operating System Kernels}},
    howpublished = {Poster at the USENIX Symposium on Operating Systems Design and Implementation (OSDI)},
    date = {2014-10-06/2014-10-08},
    year = {2014},
    address = {Broomfield, CO, USA},
    url = {http://orbit.dtu.dk/services/downloadRegister/104321268/poster_A0.pdf}
    }

2013

  • Andrea Cavallaro and Fabio Poiesi. Are Smart Cameras Ready for Prime Time?. Industrial Panel Discussion at the 10th IEEE International Conference on Advanced Video and Signal-Based Surveillance (AVSS 2013), Krakow, Poland, 2013.
    [BibTeX] [Download handouts]
    @Misc{2013-08-CAVALLARO,
    author = {Andrea Cavallaro and Fabio Poiesi},
    date = {2013-08-27},
    title = {{Are Smart Cameras Ready for Prime Time?}},
    howpublished = {Industrial Panel Discussion at the 10th IEEE International Conference on Advanced Video and Signal-Based Surveillance (AVSS 2013), Krakow, Poland},
    handouts = {http://www.eecs.qmul.ac.uk/~andrea/dwnld/2013.08.26_Panel_IEEE_AVSS_Cavallaro_Fabre_Attwood_Szczuko_Akgun.pdf},
    year = {2013}
    }

Patents

2016

  • Andrzej Czyzewski. Hanging Intelligent Road Sign. Patent (W.125159), Gdansk University of Technology, Poland, 2016.
    [BibTeX]
    @Misc{2016-05-CZYZEWSKI,
    author = {Andrzej Czyzewski},
    title = {{Hanging Intelligent Road Sign}},
    date = {2016-05},
    year = {2016},
    howpublished = {Patent (W.125159), Gdansk University of Technology, Poland}
    }

  • Andrzej Czyzewski and Józef Kotus. Sound Intensity Probe. Patent (W.125097), Gdansk University of Technology, Poland, 2016.
    [BibTeX]
    @Misc{2016-05-KOTUS,
    author = {Andrzej Czyzewski and J\'{o}zef Kotus},
    title = {{Sound Intensity Probe}},
    date = {2016-05},
    year = {2016},
    howpublished = {Patent (W.125097), Gdansk University of Technology, Poland}
    }

2015

  • Eugenio Villar Bonet, Patricia Maria Martínez Mediavilla, Francisco José Alcalà Galàn, Pablo Pedro Sànchez Espeso, and Víctor Fernàndez Solorzano. Método y sistema de localización espacial mediante marcadores luminosos para cualquier ambiente. Patent (ES-2543038-B2), Universidad de Cantabria, Spain, 2015.
    [BibTeX]
    @Misc{2015-08-VILLAR,
    author = {Eugenio {Villar Bonet} and Patricia Maria {Mart\'{i}nez Mediavilla} and Francisco Jos\'{e} {Alcal\`{a} Gal\`{a}n} and Pablo Pedro {S\`{a}nchez Espeso} and V\'{i}ctor {Fern\`{a}ndez Solorzano}},
    title = {{M\'{e}todo y sistema de localizaci\'{o}n espacial mediante marcadores luminosos para cualquier ambiente}},
    date = {2015-08-13},
    year = {2015},
    howpublished = {Patent (ES-2543038-B2), Universidad de Cantabria, Spain}
    }

2014

  • Janusz Cichowski and Andrzej Czyzewski. System for monitoring objects, especially goods in the trade facilities. Patent (P.407474), Gdansk University of Technology, Poland, 2014.
    [BibTeX]
    @Misc{2014-03-CICHOWSKI,
    author = {Janusz Cichowski and Andrzej Czyzewski},
    title = {{System for monitoring objects, especially goods in the trade facilities}},
    date = {2014-03},
    year = {2014},
    howpublished = {Patent (P.407474), Gdansk University of Technology, Poland}
    }

  • Krzysztof Nyka, Lukasz Kulas, and Przemyslaw Woźnica. Sposób lokalizacji osób ilub obiektów wewnatrz ilub na zewnatrz budynków droga komunikacji optycznej i radiowej, zwłaszcza w warunkach wystepowania zakłóceń optycznych oraz układ do realizacji sposobu. Patent (P.408187), Gdansk University of Technology, Poland, 2014.
    [BibTeX]
    @Misc{2014-05-NYKA,
    author = {Krzysztof Nyka and Lukasz Kulas and Przemyslaw Wo\'{z}nica},
    title = {{Spos\'{o}b lokalizacji os\'{o}b ilub obiekt\'{o}w wewnatrz ilub na zewnatrz budynk\'{o}w droga komunikacji optycznej i radiowej, zw\l{}aszcza w warunkach wystepowania zak\l{}\'{o}ce\'{n} optycznych oraz uk\l{}ad do realizacji sposobu}},
    date = {2014-05-12},
    year = {2014},
    howpublished = {Patent (P.408187), Gdansk University of Technology, Poland}
    }

Software

2016

  • Fabio Poiesi and Andrea Cavallaro. Detection of fast incoming objects with a moving camera. 2016. Software of Detection of fast incoming objects with a moving camera
    [BibTeX] [Download software]
    @Misc{2016-09-POIESIb,
    author = {Fabio Poiesi and Andrea Cavallaro},
    title = {{Detection of fast incoming objects with a moving camera}},
    note = {Software of Detection of fast incoming objects with a moving camera},
    date = {2016-09-07},
    year = {2016},
    software = {http://www.eecs.qmul.ac.uk/~andrea/avoidance.html}
    }

2014

  • Tahir Nawaz, Fabio Poiesi, and Andrea Cavallaro. Measures of effective video tracking. 2014. Open source software
    [BibTeX] [Download software]
    @Misc{2014-01-NAWAZ-2,
    title = {Measures of effective video tracking},
    software = {http://www.eecs.qmul.ac.uk/~andrea/mtte.html},
    author = {Tahir Nawaz and Fabio Poiesi and Andrea Cavallaro},
    organization = {{Queen Mary University of London}},
    note = {Open source software},
    date = {2014-01},
    year = {2014}
    }

  • Fabio Poiesi and Andrea Cavallaro. Tracking Multiple High-Density Homogeneous Targets. 2014. Open source software
    [BibTeX] [Download software]
    @Misc{2014-07-POIESI-1,
    title = {{Tracking Multiple High-Density Homogeneous Targets}},
    software = {http://www.eecs.qmul.ac.uk/~andrea/thdt.html},
    author = {Fabio Poiesi and Andrea Cavallaro},
    organization = {{Queen Mary University of London}},
    note = {Open source software},
    date = {2014-07-24},
    year = {2014}
    }

2013

  • Thierry Lepley. KernelGenius: a tool to describe image processing graphs in a natural way for algorithm developers and optimize them for parallel architectures. 2013. Open source software
    [BibTeX] [Download software]
    @Misc{2013-10-LEPLEY,
    title = {{KernelGenius:} a tool to describe image processing graphs in a natural way for algorithm developers and optimize them for parallel architectures},
    software = {https://github.com/tlepley/KernelGenius},
    author = {Thierry Lepley},
    organization = {{STMicroelectronics}},
    note = {Open source software},
    date = {2013-10-24},
    year = {2013}
    }

Videos

2016

  • Ricardo Sanchez-Matilla, Fabio Poiesi, and Andrea Cavallaro. Online multi-target tracking with strong and weak detections. 2016. Online video, illustration of Online multi-target tracking with strong and weak detections
    [BibTeX] [Download PDF]
    @Misc{2016-09-MATILLA,
    author = {Ricardo Sanchez-Matilla and Fabio Poiesi and Andrea Cavallaro},
    title = {{Online multi-target tracking with strong and weak detections}},
    note = {Online video, illustration of Online multi-target tracking with strong and weak detections},
    date = {2016-09-07},
    year = {2016},
    url = {http://www.eecs.qmul.ac.uk/~andrea/eamtt.html}
    }

  • Fabio Poiesi and Andrea Cavallaro. Detection of fast incoming objects with a moving camera. 2016. Online video, illustration of Detection of fast incoming objects with a moving camera
    [BibTeX] [Watch video]
    @misc{2016-09-POIESIa,
    author = {Fabio Poiesi and Andrea Cavallaro},
    title = {{Detection of fast incoming objects with a moving camera}},
    note = {Online video, illustration of Detection of fast incoming objects with a moving camera},
    date = {2016-09-07},
    year = {2016},
    video = {http://www.eecs.qmul.ac.uk/~andrea/avoidance.html}
    }

2014

  • Tahir Nawaz, Fabio Poiesi, and Andra Cavallaro. Measures of Effective Video Tracking. 2014. Online video
    [BibTeX] [Watch video]
    @Misc{2014-11-NAWAZ,
    author = {Tahir Nawaz and Fabio Poiesi and Andra Cavallaro},
    title = {{Measures of Effective Video Tracking}},
    note = {Online video},
    date = {2014-11-18},
    video = {https://www.youtube.com/watch?v = pp8HQEQ-INU},
    year = {2014}
    }

  • Fabio Poiesi and Andrea Cavallaro. Tracking Multiple High-Density Homogeneous Targets. 2014. Online video
    [BibTeX] [Watch video]
    @Misc{2014-11-POIESI,
    author = {Fabio Poiesi and Andrea Cavallaro},
    title = {{Tracking Multiple High-Density Homogeneous Targets}},
    note = {Online video},
    date = {2014-11-23},
    video = {https://www.youtube.com/watch?v = zNuabwxDBHk},
    year = {2014}
    }

Comments are closed.