Physics-informed Machine Learning for Robust Pedestrian Detection in Embedded Low-power System

Thursday 26 Oct 2023, 13:00 → 14:00 UTC

Abstract:
For urban and metropolitan areas, population growth is expected to continue in the coming years. Cities and municipalities will continue to strive to provide residents with a sense of security and freedom. Camera-based surveillance systems are currently the most mature solution for urban surveillance, for example in monitoring road traffic and, or course, detecting abnormal pedestrian motion and activity at road side. However, most existing and established algorithms and systems for urban surveillance require expensive and power-hungry computer hardware. Neuromorphic sensors and computing systems are a promising alternative. The key advantages of such systems are high energy efficiency, fast and representative (relevant) data acquisition, fast local processing, improved data/identity protection, and rational budgeted use of resources. At town-scale this can provide high benefits. In this talk, we will focus on a novel and efficient real-time machine learning algorithm infused with physical motion models that is able to accurately detect and track pedestrians and bikers both day and night in urban scenarios.

About the webinar series:

This bi-monthly seminar series explores real-world applications of physics-informed machine learning (Φ-ML) methods to the engineering practice. They cover a wide range of topics, offering a cross-sectional view of the state of the art on Φ-ML research, worldwide.  

Participants have the opportunity to hear from leading researchers and learn about the latest developments in this emerging field. These seminars also offer the chance to identify and spark collaboration opportunities.

More info / subscribe: https://www.turing.ac.uk/events/phi-ml-meets-engineering-physics-informed-machine-learning-robust-pedestrian-detection