Year: 2025
Type: preprint
Abstract: Human mesh recovery (HMR) from a single image is inherently ill-posed due to depth ambiguity and occlusions. Probabilistic methods have tried to solve this by generating numerous plausible 3D human mesh predictions, but they often exhibit misalignment with 2D image observations and weak robustness to in-the-wild images. To address these issues, we propose ADHMR, a framework that Aligns a Diffusion-based HMR model in a preference optimization manner. First, we train a human mesh prediction assessment model, HMR-Scorer, capable of evaluating predictions even for in-the-wild images without 3D annotations. We then use HMR-Scorer to create a preference dataset, where each input image has a pair of winner and loser mesh predictions. This dataset is used to finetune the base model using direct preference optimization. Moreover, HMR-Scorer also helps improve existing HMR models by data cleaning, even with fewer training samples. Extensive experiments show that ADHMR outperforms current state-of-the-art methods. Code is available at: https://github.com/shenwenhao01/ADHMR.
Source: arXiv (Cornell University)
Authors: Shen, Wenhao, Yin, Wanqi, Yang, Xiaofeng, Chen, Cheng, Song, Chaoyue, Cai, Zhongang, Yang, Lei, Wang, Hao, Lin, Guosheng
Language: English
Subfield: Computational Mechanics
Field: Engineering
Domain: Physical Sciences
Open Access status: green
Funder: Ministry of Education, India
OpenAlex ID: W4416445243
Cites:
Cited by:
Related to: