Embedded Vision Systems Group, Computer Vision Laboratory AGH

The Computer Vision Laboratory is a part of Department of Automatic Control and Robotics, Faculty of Electrical Engineering, Automatics, Computer Science and Biomedical Engineering, AGH University of Science and Technology, Kraków, Poland. We conduct research in the following areas: - real-time image and video stream processing, analysis and recognition using reprogrammable devices (FPGAs) and programmable GPUs (basic algorithms like foreground object segmentation, moving object segmentation, object detection, classification and tracking as well as advanced video surveillance systems), - embedded vision systems implemented on platforms with low energy consumption, i.e. FPGA, heterogeneous programmable SoC (e.g. Zynq SoC) and embedded GPU, - control of autonomous unmanned aerial vehicles (drones) and cars based on visual feedback and data from a LiDAR sensor, - deep convolutional neural networks with limited precision (quantized neural networks - QNN), - event cameras and splining neural networks (i.e. neuromorphic computing), - deep neural networks in early detection of melanomas; computer-aided dermatopathology, - anomaly detection with the use of pre-trained CNN architectures, - detection and analysis of patterns e.g. cell detection, - reconstructing images' missing areas with generative models, - bacteria response clustering. Our teaching activities are focused mainly on the Automatic Control and Robotics course (bachelor and masters) from introduction to programming languages (Python, C, C++), signal processing, computer vision and machine learning to advanced subjects on advanced and embedded vision systems, deep neural networks, medical image processing, autonomous vehicles, OpenCL, CUDA and hardware-software vision systems.