Reinforcement learning-based motion imitation for physiologically plausible musculoskeletal motor control

2025-03-18

AI Summary

Understanding neuromuscular control in musculoskeletal movement remains challenging due to Bernstein’s redundancy problem and the lack of interpretable, biologically grounded computational models. Method: We propose KINESIS—a model-free reinforcement learning framework for motion imitation on a high-fidelity, full-muscle-driven lower-limb musculoskeletal model (80 muscles, 20 degrees of freedom). It integrates Proximal Policy Optimization (PPO), physiologically constrained muscle dynamics, text-to-motion transfer leveraging a pretrained T2M model, and EMG-based interpretability evaluation. Contribution/Results: KINESIS achieves the first EMG-interpretable motion imitation (r > 0.7) on such a high-fidelity musculoskeletal model, providing computational neuroscience evidence for resolving motor redundancy. It reproduces naturalistic motions with only 1.9 hours of motion-capture data. Furthermore, it supports natural-language instruction conditioning and goal-directed fine-tuning, thereby bridging computational motor control theory and experimental motor neuroscience.

Technology Category

• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Intelligent Robots: Behavior Learning & Control

Application Category

• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

How do humans move? The quest to understand human motion has broad applications in numerous fields, ranging from computer animation and motion synthesis to neuroscience, human prosthetics and rehabilitation. Although advances in reinforcement learning (RL) have produced impressive results in capturing human motion using simplified humanoids, controlling physiologically accurate models of the body remains an open challenge. In this work, we present a model-free motion imitation framework (KINESIS) to advance the understanding of muscle-based motor control. Using a musculoskeletal model of the lower body with 80 muscle actuators and 20 DoF, we demonstrate that KINESIS achieves strong imitation performance on 1.9 hours of motion capture data, is controllable by natural language through pre-trained text-to-motion generative models, and can be fine-tuned to carry out high-level tasks such as target goal reaching. Importantly, KINESIS generates muscle activity patterns that correlate well with human EMG activity. The physiological plausibility makes KINESIS a promising model for tackling challenging problems in human motor control theory, which we highlight by investigating Bernstein's redundancy problem in the context of locomotion. Code, videos and benchmarks will be available at https://github.com/amathislab/Kinesis.

Problem

• Develops a reinforcement learning framework for musculoskeletal motor control.
• Achieves physiologically plausible muscle activity patterns in motion imitation.
• Addresses Bernstein's redundancy problem in human locomotion.

Innovation

• Reinforcement learning for musculoskeletal motor control
• Model-free motion imitation framework (KINESIS)
• Controllable by natural language text-to-motion models