With its countless advantages, vision-based knowledge inference is one of the most effective ways to understand our surroundings. Exploration of distant planets and galaxies, understanding nano creatures, the perception of the environment in everyday activities such as driving and playing, behavioral understanding—these are all based on vision. While human-made systems are still in early stages of the vision-based development, biological systems have developed over general very powerful and integrated mechanisms, capable of inferring their surroundings in a fraction of second.
All-a.round vision discusses vision-based perception of the world around us, inspired by biological systems to devise computing infrastructure that fulfills SWAP constraints. We explore the versatility of system architectures and components, approaches, as well as the perspective view and its advantage over conventional single view. Applications are used as motivation to drive the development of the technology.
This seminar will address the broad research topic of smart grids’ cyber-physical security. This talk will present a cross-layered framework that focuses on detection, identification, and reaction, specifically on enhancing system robustness to contingencies through distributed control and resiliency to cyber-attacks through machine learning and state estimation, which enables anomaly detection and self-healing communication networks.