HRI

We introduce Self-Constrained Reasoning, a novel approach that embeds an LTL-specific synchronous grammar into hierarchical reinforcement learning, enabling LLM-based task specification while preserving formal validity and generalization.

This paper presents an optimization-based control framework for powered knee prostheses to perform stair ascent.

This paper presents a novel framework for powered knee prostheses to achieve obstacle avoidance by directly integrating environmental information.

We develop a model-free reinforcement learning framework that enables the prosthesis to adapt to diverse human movement patterns through cooperative policy learning.