This paper focuses on a critical challenge in robotics: translating text-driven human motions into executable actions for humanoid robots, enabling efficient and cost-effective learning of new behaviors. While existing text-to-motion generation methods achieve semantic alignment between language and motion, they often produce kinematically or physically infeasible motions unsuitable for real-world deployment. To bridge this sim-to-real gap, we propose Reinforcement Learning from Physical Feedback (RLPF), a novel framework that integrates physics-aware motion evaluation with text-conditioned motion generation. RLPF employs a motion tracking policy to assess feasibility in a physics simulator, generating rewards for fine-tuning the motion generator. Furthermore, RLPF introduces an alignment verification module to preserve semantic fidelity to text instructions. This joint optimization ensures both physical plausibility and instruction alignment. Extensive experiments show that RLPF greatly outperforms baseline methods in generating physically feasible motions while maintaining semantic correspondence with text instruction, enabling successful deployment on real humanoid robots.

In this paper, we present Reinforcement Learning from Physical Feedback (RLPF), a novel framework that resolves physical inconsistency in motion generation models for humanoid robots. RLPF integrates two key components, a pretrained motion tracking policy and an alignment verification module, which ensures that generated motions are both physically feasible and semantically correspond with the input instructions. Extensive experiments demonstrate the effectiveness of our RLPF approach, showing a pathway for the humanoid robots in real-world applications.