P U L S A R
Stars is the successor of Orion. This research project-team focuses on Activity Recognition. More precisely we are interested in the real-time semantic interpretation of dynamic scenes observed by sensors. Thus, we study long-term saptio-temporal activities performed by human beings, animals or vehicles in the physical world. The major issue in semantic interpretation of dynamic scenes is the gap between the subjective interpretation of data and the objective measures provided by the sensors. Our approach, in order to address this problem, is to keep a clear boundary between the application dependant subjective interpretations and the objective general analysis of the videos. Pulsar propose new techniques in the field of cognitive vision and cognitive systems for physical object recognition, activity uderstanding, activity learning, system design and evaluation and focus on two main application domains: safety/security and healthcare.
We have two main research themes:
- Scene understanding for activity recognition: scene understanding aims at solving the complete interpretation problem ranging from low level signal analysis to semantic description of what is happening in the scene viewed by video cameras and possibly other sensors. We work more particularly on perception, understanding and learning.
- Software architecture for activity recognition: this research direction consists in studying generic systems for activity recognition and in elaborating a methodology for their design. We wish to ensure genericity, modularity, re usability, extensibility, dependability, and maintainability. We work more particularly on models, platform architecture, and system safeness.