Saved in:
Bibliographic Details
Main Authors: Yao, Nanjie, Ren, Junlong, Shen, Wenhao, Wang, Hao
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.05680
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909034963009536
author Yao, Nanjie
Ren, Junlong
Shen, Wenhao
Wang, Hao
author_facet Yao, Nanjie
Ren, Junlong
Shen, Wenhao
Wang, Hao
contents 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
format Preprint
id arxiv_https___arxiv_org_abs_2605_05680
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle MotionGRPO: Overcoming Low Intra-Group Diversity in GRPO-Based Egocentric Motion Recovery
Yao, Nanjie
Ren, Junlong
Shen, Wenhao
Wang, Hao
Computer Vision and Pattern Recognition
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
title MotionGRPO: Overcoming Low Intra-Group Diversity in GRPO-Based Egocentric Motion Recovery
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.05680