09:00-09:50 | General Lecture II
Chair: Arkadiusz Or³owski | 09:50-10:00 | Coffee break | 10:00-11:20 | Session III. Biomedical applications
Chair: Joanna Jaworek-Korjakowska
1 Software for CT-image analysis to assist the choice of mechanical-ventilation settings in acute respiratory distress syndrome Eduardo E. Dįvila Serrano CNRS, Franēois Dhelft Univ. Lyon 1 - Hospices Civils de Lyon, Laurent Bitker Univ. Lyon 1 - Hospices Civils de Lyon, Jean-Christophe Richard Univ. Lyon 1 - Hospices Civils de Lyon, Maciej Orkisz Univ. Lyon 1, CREATIS, France |  show paper details | 
2 Tuberculosis Abnormality Detection in Chest X-Rays: A Deep Learning Approach Mustapha Oloko-Oba, Serestina Viriri |  show paper details | 
3 On the Effect of DCE MRI Slice Thickness and Noise on Estimated Pharmacokinetic Biomarkers - A Simulation Study Jakub Jurek, Lars A. Reisæter, Marek Kociński, Andrzej Materka |  show paper details | 
4 Nuclei detection with local threshold processing in DAB&H stained breast cancer biopsy images Lukasz Roszkowiak (IBIB PAN), Jakub Zak (IBIB PAN), Krzysztof Siemion (IBIB PAN), Dorota Pijanowska (IBIB PAN), Anna Korzynska (IBIB PAN) |  show paper details |
| 10:20-13:00 | Lunch break | 13:00-14:20 | Session IV. 3D vision
Chair: Andrzej ¦luzek
1 Performance Evaluation of Selected 3D Keypoint Detector-Descriptor Combinations Paula Stancelova, Elena Sikudova, Zuzana Cernekova |  show paper details | 
2 A vision based hardware-software real-time control system for the autonomous landing of an UAV Krzysztof Blachut, Hubert Szolc, Mateusz Wasala, Tomasz Kryjak, Marek Gorgon, AGH University of Science and Technology, Krakow, Poland |  show paper details | 
3 RGB-D and Lidar Calibration Supported by GPU Artur Wilkowski, Warsaw University of Technology and Dariusz Mańkowski, United Robots |  show paper details | 
4 Optimisation of a Siamese Neural Network for Real-Time Energy Efficient Object Tracking Dominika Przewlocka, Mateusz Wasala, Hubert Szolc, Krzysztof Blachut, Tomasz Kryjak - AGH University of Science and Technology, Krakow, PolandSiamese neural networks , DNN, QNN, object tracking, FPGA, embedded vision system In this paper the research on optimisation of visual object tracking using a Siamese neural network for embedded vision systems is presented. It was assumed that the solution shall operate in real-time, preferably for a high resolution video stream, with the lowest possible energy consumption. To meet these requirements, techniques such as the reduction of computational precision and pruning were considered. Brevitas, a~tool dedicated for optimisation and quantisation of neural networks for FPGA implementation, was used. A~number of training scenarios were tested with varying levels of optimisations - from integer uniform quantisation with 16 bits to ternary and binary networks. Next, the influence of these optimisations on the tracking performance was evaluated. It was possible to reduce the size of the convolutional filters up to 10 times in relation to the original network. The obtained results indicate that using quantisation can significantly reduce the memory and computational complexity of the proposed network while still enabling precise tracking, thus allow to use it in embedded vision systems. Moreover, quantisation of weights positively affects the network training by decreasing overfitting. |  hide paper details |
|
|