Description
Studying how humans control muscles to walk or run is hard with simulations that ignore physiology. Kinesis from EPFL uses reinforcement learning to make musculoskeletal models imitate human motion with physiologically plausible control.
It predicts EMG and proprioception for movement science and rehabilitation research; published at ICRA 2026.
Imitation:Human motion.
Musculoskeletal:Physiologically real.
EMG:Predicted.
It predicts EMG and proprioception for movement science and rehabilitation research; published at ICRA 2026.
Features
Imitation:Human motion.
Musculoskeletal:Physiologically real.
EMG:Predicted.
