FairHuman: Boosting Hand and Face Quality in Human Image Generation with Minimum Potential Delay Fairness in Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Yuxuan, Cao, Tianwei, Zhang, Huayu, He, Zhongjiang, Liang, Kongming, Ma, Zhanyu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912462667776000
author Wang, Yuxuan
Cao, Tianwei
Zhang, Huayu
He, Zhongjiang
Liang, Kongming
Ma, Zhanyu
author_facet Wang, Yuxuan
Cao, Tianwei
Zhang, Huayu
He, Zhongjiang
Liang, Kongming
Ma, Zhanyu
contents Image generation has achieved remarkable progress with the development of large-scale text-to-image models, especially diffusion-based models. However, generating human images with plausible details, such as faces or hands, remains challenging due to insufficient supervision of local regions during training. To address this issue, we propose FairHuman, a multi-objective fine-tuning approach designed to enhance both global and local generation quality fairly. Specifically, we first construct three learning objectives: a global objective derived from the default diffusion objective function and two local objectives for hands and faces based on pre-annotated positional priors. Subsequently, we derive the optimal parameter updating strategy under the guidance of the Minimum Potential Delay (MPD) criterion, thereby attaining fairness-ware optimization for this multi-objective problem. Based on this, our proposed method can achieve significant improvements in generating challenging local details while maintaining overall quality. Extensive experiments showcase the effectiveness of our method in improving the performance of human image generation under different scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2507_02714
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FairHuman: Boosting Hand and Face Quality in Human Image Generation with Minimum Potential Delay Fairness in Diffusion Models
Wang, Yuxuan
Cao, Tianwei
Zhang, Huayu
He, Zhongjiang
Liang, Kongming
Ma, Zhanyu
Computer Vision and Pattern Recognition
Artificial Intelligence
Image generation has achieved remarkable progress with the development of large-scale text-to-image models, especially diffusion-based models. However, generating human images with plausible details, such as faces or hands, remains challenging due to insufficient supervision of local regions during training. To address this issue, we propose FairHuman, a multi-objective fine-tuning approach designed to enhance both global and local generation quality fairly. Specifically, we first construct three learning objectives: a global objective derived from the default diffusion objective function and two local objectives for hands and faces based on pre-annotated positional priors. Subsequently, we derive the optimal parameter updating strategy under the guidance of the Minimum Potential Delay (MPD) criterion, thereby attaining fairness-ware optimization for this multi-objective problem. Based on this, our proposed method can achieve significant improvements in generating challenging local details while maintaining overall quality. Extensive experiments showcase the effectiveness of our method in improving the performance of human image generation under different scenarios.
title FairHuman: Boosting Hand and Face Quality in Human Image Generation with Minimum Potential Delay Fairness in Diffusion Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2507.02714