GenColor: Generative and Expressive Color Enhancement with Pixel-Perfect Texture Preservation

NeurIPS 2025 Spotlight

Yi Dong1†, Yuxi Wang1†, Xianhui Lin2, Wenqi Ouyang1, Zhiqi Shen1*, Peiran Ren2*, Ruoxi Fan1, Rynson W. H. Lau3
1Nanyang Technological University, 2Alibaba Group, 3City University of Hong Kong
†Equal contribution, *Corresponding authors
Paper Dataset Code
Teaser

GenColor achieves expressive color enhancement with superior texture preservation.

Abstract

Color enhancement is a crucial yet challenging task in digital photography. It demands methods that are (i) expressive enough for fine-grained adjustments, (ii) adaptable to diverse inputs, and (iii) able to preserve texture. Existing approaches typically fall short in at least one of these aspects, yielding unsatisfactory results.

We propose GenColor, a novel diffusion-based framework for sophisticated, texture-preserving color enhancement. GenColor reframes the task as conditional image generation. Leveraging ControlNet and a tailored training scheme, it learns advanced color transformations that adapt to diverse lighting and content. We train GenColor on ARTISAN, our newly collected large-scale dataset of 1.2M high-quality photographs specifically curated for enhancement tasks.

To overcome texture preservation limitations inherent in diffusion models, we introduce a color transfer network with a novel degradation scheme that simulates texture–color relationships. This network achieves pixel-perfect texture preservation while enabling fine-grained color matching with the diffusion-generated reference images. Extensive experiments show that GenColor produces visually compelling results comparable to those of expert colorists and surpasses state-of-the-art methods in both subjective and objective evaluations.

Method

Method Overview

GenColor addresses three key challenges through a three-phase process:

  1. Generative Color Enhancement: A diffusion-based Color Generation module creates expressive color enhancements conditioned on the input image, trained on the ARTISAN-1M dataset.
  2. Texture Preservation: A Texture Preservation module maintains texture integrity while transferring colors from the color reference to the input image.
  3. Global Adjustment: A Global Adjustment module applies lightweight filters to enhance contrast and saturation.

Supplementary Video

Comparisons

Click on the buttons to compare different methods against the Input and GenColor.

Texture Preservation in Commercial Models

Even the latest commercial closed-source models still have texture change problems. GenColor can effectively help optimize their results, achieving pixel-perfect texture preservation.

Click on the buttons to compare different commercial models.

BibTeX

@inproceedings{dong2025gencolor,
  title={GenColor: Generative and Expressive Color Enhancement with Pixel-Perfect Texture Preservation},
  author={Dong, Yi and Wang, Yuxi and Lin, Xianhui and Ouyang, Wenqi and Shen, Zhiqi and Ren, Peiran and Fan, Ruoxi and Lau, Rynson W. H.},
  booktitle={NeurIPS},
  year={2025}
}