AiMIND
AI4AD
* Flagship
AI for Autonomous Driving
End-to-end autonomous driving, motion planning, state estimation, trajectory prediction, and control research.
- End-to-end learning for perception and decision-making in autonomous driving
- Motion planning under uncertainty and interaction-aware driving conditions
- Uncertainty-aware state estimation for robust perception and control under dynamic driving conditions
- Trajectory prediction for surrounding agents and long-horizon safety
- Control algorithms that integrate prediction and decision feedback
AI for Human-Centered Mobility
Risk-aware voice assistance systems that fuse in-vehicle and surround video, voice, and CAN data to detect hazardous situations and guide drivers.
- Building a systematic dataset preprocessing framework
- Designing multi-modal AI models that combine voice, vehicle CAN data, and camera input
- Designing priority decision logic for driver risk detection and warnings
- Model quantization and optimization for in-vehicle deployment
AI for Reliable Chemical Sensing
Machine learning methods for robust, noise-tolerant chemical and environmental sensing.
- Gas classification & concentration estimation
- Deep ensemble-based uncertainty quantification
- Reliable, AI-enabled chemical sensing
AI for Battery Management Systems
Simulation-to-real reinforcement learning for battery state-of-health estimation and autonomous optimal charging.
- Battery SOH & uncertainty estimation
- Reinforcement learning-based adaptive charging
- Simulation-to-real policy optimization