Context
- WSNs for the smart city
- Autonomous vehicles
- Portable and confidential activity recognition
- Sensory substitution for the visually impaired
- space applications
Objectives
- Study and draw inspiration from the biological brain to explore new forms of artificial intelligence
- Make AI more efficient and autonomous
- Reduce the energy consumption of these complex algorithms
- Make IA algorithms compatible with real-time execution constraints
Challenges
- Optimize new models of artificial neural networks to integrate them into embedded electronic circuits (microcontrollers, FPGA, eFPGA, ASIC...)
- Confronting AI with biological intelligence
- Jointly optimize the compression and robustness of neural networks to make them applicable in realistic environments
Current projects
Integrating embedded neural networks and
self-mixing interferometry for smart sensors design
Past projects
