Paint-Anything enables precise color control for image generation and editing
Paint-Anything: Unified Any-Color Control for Image Generation and Editing
Computer Vision and Pattern RecognitionArtificial IntelligenceMachine Learning
Summary
Changing the exact color of objects in computer-generated images is hard, especially when you want to pick any color from the full digital palette. The authors created Paint-Anything, a system that learns to associate specific color codes (hex values) with objects in images to generate or edit images with precise color control. They built a large dataset from real pictures mixed with exact color samples to train their model and developed tests to measure how well it matches targeted colors. This approach improves color accuracy in both creating new images and editing existing ones.
What this means in practice
- •For graphic designers: Generate and edit images with any specified exact color for object elements, enhancing precise creative workflows.
- •For game developers: Apply exact color modifications to in-game assets during design or updates without needing specialized color tuning tools.
Authors
Ji Xie, Dewei Zhou, Xinyu Huang, Zhennan Chen, Xun Wang
Abstract
Professional design requires any-color control: the ability to specify an object's target color with any 24-bit hex value for image generation and editing. Prior work has explored color generation, editing, and colorization, but often relies on dedicated color representations or specialized inference procedures. Advances in large language models offer a simpler starting point: even compact models can associate hex values with color semantics. We present Paint-Anything, which learns a shared hex-prompt interface for generation and editing through object-level color supervision. We develop a data pipeline that constructs Paint-500K from real images through object grounding, perceptual color labeling, and editing-pair synthesis. Since shadows make real-image labels only approximate colors, we complement this supervision with pure-color anchors whose pixels exactly match their paired hex values. These anchors are used only at high-noise timesteps, leaving low-noise training to natural images. We further introduce Any Color Benchmark (ACBench), comprising ACBench-T2I and ACBench-Edit, to measure object-level hex color fidelity across both tasks. On FLUX.2-4B, Paint-Anything improves ACBench-T2I and ACBench-Edit scores by 85.3% and 28.3%, respectively, relative to the base model, with ablations supporting the training recipe. It also achieves the highest average CompColor score among the compared methods.