Understanding the driver before risk becomes an event.
This system monitors facial geometry and head orientation to identify patterns associated with drowsiness and distraction. The goal is to turn perception outputs into timely, reliable intervention signals in a vehicle environment.
The product work also includes communication through the OBD-II port and transfer to a central server, connecting the neural perception pipeline with the wider vehicle and monitoring architecture.
Key contributions
- Implemented neural face-mesh and head-orientation estimation for real-time use.
- Developed decision logic for drowsiness and distraction detection.
- Connected perception events with vehicle data over OBD-II.
- Supported central-server communication for remote monitoring and analysis.
