DiffBody: Human Body Restoration by Imagining with Generative Diffusion Prior

TL;DR

DiffBody introduces a human body-aware diffusion model that leverages domain-specific knowledge, attention modules, and text prompts to significantly improve human body restoration quality over existing methods.

Contribution

The paper presents a novel diffusion-based approach with domain-specific enhancements, including a body attention module and text prompt integration, for superior human body restoration.

Findings

Outperforms existing methods quantitatively and qualitatively

Effectively restores surface textures, accessories, and limb structures

Provides a new benchmark dataset for human body restoration

Abstract

Human body restoration plays a vital role in various applications related to the human body. Despite recent advances in general image restoration using generative models, their performance in human body restoration remains mediocre, often resulting in foreground and background blending, over-smoothing surface textures, missing accessories, and distorted limbs. Addressing these challenges, we propose a novel approach by constructing a human body-aware diffusion model that leverages domain-specific knowledge to enhance performance. Specifically, we employ a pretrained body attention module to guide the diffusion model's focus on the foreground, addressing issues caused by blending between the subject and background. We also demonstrate the value of revisiting the language modality of the diffusion model in restoration tasks by seamlessly incorporating text prompt to improve the quality of surface texture and additional clothing and accessories details. Additionally, we introduce a diffusion sampler tailored for fine-grained human body parts, utilizing local semantic information to rectify limb distortions. Lastly, we collect a comprehensive dataset for benchmarking and advancing the field of human body restoration. Extensive experimental validation showcases the superiority of our approach, both quantitatively and qualitatively, over existing methods.