Wearable & Healthcare
We develop AI-driven approaches for continuous health monitoring and assistive robotics. By combining wearable sensor data (IMU, EMG, pressure insoles) with foundation models, we enable real-time estimation of body kinematics, kinetics, and activity states across everyday settings. These capabilities power adaptive control of exoskeletons and wearable robotic systems — interpreting user intent and delivering coordinated assistance for rehabilitation and mobility enhancement. Our world models learn to predict human movement dynamics, enabling proactive and personalized interventions.
Intention Detection for Wearable Robot Hand
Ankle Exosuit for Post-Stroke Rehabilitation
Driver State Prediction for Takeover
Pose Estimation using Sparse IMUs
Humanoid IMU Teleoperation
Key Research Topics
- Foundation Models for Medical and Healthcare Applications
- World Models for Human Movement Prediction
- Human Kinematics and Kinetics Estimation
- Intention Detection and Activity Recognition
- Exoskeletons and Wearable Robotic Systems
- Wearable Sensors for Rehabilitation and Digital Healthcare
