arXiv：cs.AI· Vincent Auriau, Emmanuel Malherbe, Matthieu Perrot·· 16 小时前
基于弱分割引导的 GAN 实现逼真图像色彩编辑
Weak Segmentation-Guided GAN for~Realistic Color Edition
AI 导读
研究者提出一种利用任意可微分分割模型的 GAN 架构，用于逼真的图像颜色编辑。该框架在多个用例上取得 SOTA 结果，兼顾生成无伪影的逼真图像与精确的颜色控制能力，克服了传统分割加直方图匹配方法牺牲真实性、而 GAN 难以控制的难题。
正文
Abstract:Editing the color of images in a realistic way finds many applications such as changing the perception of an image, data augmentation or film post processing. The design of an automatic tool is a complex and long addressed challenge. In particular, two properties are difficult to meet altogether: generating realistic results without artifacts and the possibility to precisely choose the future color. Conventional methods using segmentation and histogram matching maximize the controllability but also introduce a lack of realism or are complex to automate. On the contrary, GANs that specialize in realism are difficult to control. To overcome these challenges, we propose a novel GAN architecture leveraging any differentiable segmentation model. We demonstrate the genericness of our framework that presents state of the art results on different use cases. It generates images that look realistic while offering a precise color control.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.16620 [cs.AI] |
| (or arXiv:2610.16620v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.16620 arXiv-issued DOI via DataCite (pending registration) |
|
| Journal reference: | Image Analysis and Processing -- ICIAP 2023, 14233, Springer Nature Switzerland, pp.487-499, 2023, Lecture Notes in Computer Science |
Submission history
From: Vincent Auriau [view email] [via CCSD proxy]
[v1]
Mon, 12 Oct 2026 11:36:01 UTC (7,331 KB)
来源：arXiv：cs.AI · arxiv.org