arXiv 2604.212892026-04-23cs.CV

AttDiff-GAN：一种用于面部属性编辑的混合扩散-生成对抗网络框架

AttDiff-GAN: A Hybrid Diffusion-GAN Framework for Facial Attribute Editing

GuangDong Province Key Lab of Information Security Technology and School of Computer Science and Engineering（广东信息安全技术重点实验室和计算机科学与工程学院）；School of Cyber Science and Technology, Shenzhen Campus（深圳校区网络科学与技术学院）；Guangdong Laboratory of Machine Perception and Intelligent Computing, Faculty of Engineering（广东机器感知与智能计算实验室，工程学院）

Wenmin Huang, Weiqi Luo, Xiaochun Cao, Jiwu Huang

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出AttDiff-GAN框架，结合GAN属性操控与扩散图像生成，通过特征级对抗学习和PriorMapper、RefineExtractor提升风格-属性对齐，实现更精确的面部属性编辑。

摘要：

面部属性编辑旨在修改目标属性的同时保持无关属性内容和整体图像保真度。现有基于GAN的方法虽具有良好的可控性，但常面临风格代码与属性语义之间的对齐问题。扩散方法能生成高度逼真的图像，但其编辑精度受限于不同属性之间语义方向的交织。本文提出AttDiff-GAN，一种结合基于GAN的属性操控与基于扩散的图像生成的混合框架。关键挑战在于一步对抗学习与多步扩散去噪之间的不一致性，使有效优化困难。为解决此问题，我们通过引入特征级对抗学习方案学习显式属性操控，并利用操控后的特征指导扩散过程进行图像生成，同时去除对基于语义方向的编辑的依赖。此外，我们通过引入PriorMapper将面部先验纳入风格生成，并通过Transformer捕获全局语义关系以更精确地提取风格。在CelebA-HQ上的实验结果表明，所提方法在定性和定量评估中均实现了更准确的面部属性编辑和更好的非目标属性保留。

英文摘要：

Facial attribute editing aims to modify target attributes while preserving attribute-irrelevant content and overall image fidelity. Existing GAN-based methods provide favorable controllability, but often suffer from weak alignment between style codes and attribute semantics. Diffusion-based methods can synthesize highly realistic images; however, their editing precision is limited by the entanglement of semantic directions among different attributes. In this paper, we propose AttDiff-GAN, a hybrid framework that combines GAN-based attribute manipulation with diffusion-based image generation. A key challenge in such integration lies in the inconsistency between one-step adversarial learning and multi-step diffusion denoising, which makes effective optimization difficult. To address this issue, we decouple attribute editing from image synthesis by introducing a feature-level adversarial learning scheme to learn explicit attribute manipulation, and then using the manipulated features to guide the diffusion process for image generation, while also removing the reliance on semantic direction-based editing. Moreover, we enhance style-attribute alignment by introducing PriorMapper, which incorporates facial priors into style generation, and RefineExtractor, which captures global semantic relationships through a Transformer for more precise style extraction. Experimental results on CelebA-HQ show that the proposed method achieves more accurate facial attribute editing and better preservation of non-target attributes than state-of-the-art methods in both qualitative and quantitative evaluations.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。