MotionGRPO: Overcoming Low Intra-Group Diversity in GRPO-Based Egocentric Motion Recovery

TL;DR

MotionGRPO introduces a reinforcement learning framework with a noise-injection strategy to improve 3D human motion recovery from head-mounted device signals, addressing local reconstruction errors and intra-group diversity issues.

Contribution

It presents a novel diffusion sampling approach modeled as a Markov decision process, enhancing global and local motion reconstruction accuracy.

Findings

Achieves state-of-the-art performance in 3D motion recovery

Improves visual fidelity over existing methods

Addresses intra-group diversity issues with noise-injection

Abstract

This paper studies full-body 3D human motion recovery from head-mounted device signals. Existing diffusion-based methods often rely on global distribution matching, leading to local joint reconstruction errors. We propose MotionGRPO, a novel framework leveraging reinforcement learning post-training to inject fine-grained guidance into the diffusion process. Technically, we model diffusion sampling as a Markov decision process optimized via Group Relative Policy Optimization (GRPO). To this end, we introduce a hybrid reward mechanism that combines a learned conditioned perceptual model for global visual plausibility and explicit constraints for local joint precision. Our key technical insight is that policy optimization in diffusion-based recovery suffers from vanishing gradients due to limited intra-group sample diversity. To address this, we further introduce a noise-injection strategy that explicitly increases sample variance and stabilizes learning. Extensive experiments demonstrate that MotionGRPO achieves state-of-the-art performance with superior visual fidelity