Profile of the Signal Processing Group
Our department focuses on the research, development and implementation of advanced digital signal and image processing and AI inference algorithms, mainly in the fields of audio and ultrasound processing and scene analysis (image segmentation, motion detection, object detection). The department also develops methods in Bayesian statistics for the recursive estimation of mixtures with different component distributions, addressing the analysis of clusters with non-Gaussian characteristics. The theory is applied in practical contexts, such as traffic flow states, vehicle counts, accident analysis, and crime data, and is also taught at the Faculty of Transportation Sciences, CTU in Prague.
We work mainly with algorithms which can be implemented with fast execution, small memory footprint, small chip area and low power consumption. Our aim is to work with our project partners on implementation of these algorithms in real demonstrators and prototypes. Our target platforms are SoC devices equipped with multicore processors, Field Programmable Gate Arrays and dedicated HW accelerator engines. We are focused mostly on embedded SoC solutions based on AMD Zynq, Zynq Ultrascale+ devices (28nm, 16nm technology) and AMD Versal devices (7nm technology). We also use Matlab/Simulink framework and SW libraries to specify, model and verify algorithms which we subsequently convert and synthesize to HW by AMD high-level synthesis tools.
See presentation and videos for the Department of Signal Processing evaluation 9th March 2021:
- presentation (.pdf 4 MB)
- presentation with embedded videos (.pptx 192 MB)
- video 1 (.mp4 10 MB)
- video 2 (.mp4 87 MB)
- video 3 (.mp4 78 MB)
- video 4 (.mp4 6 MB)
Our aim is also to help industrial partners to solve implementation issues in all their complexity. The key partner has been Technical Development of SKODA AUTO a.s. where we have main competence in development of electronic systems for testing Human-Machine-Interface. This includes support for the development and testing of various levels of autonomous driving, including alternative powertrains such as fully electric vehicles (FEVs), as well as support for various forms of mobility, including urban mobility and micro-mobility.
Our aim is not only to deal with the theoretical design of algorithms but also to help industrial partners to solve implementation issues in all their complexity. The key partner is Technical Development of SKODA AUTO a.s. where we have main competence in development electronic systems for testing Human-Machine-Interface.
Our department has been participating in R&D projects supported by the Horizon Europa programme, mainly in projects of the Chips Joint Undertaking.
Among further activities belongs e.g. long-term participation in development international technical standards in Working Groups ISO, For example, ISO/IEC JTC 1/SC7 Software and Systems Engineering, cybersecurity, or unmanned systems (UAS).