Consistent Feature Transport for Image Relighting
2026-07-20 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors address the problem of changing the lighting in an image while keeping everything else, like the person's identity and shape, the same. They propose a new method called Consistent Feature Transport (CFT) that views relighting as moving lighting features from one image to another in a consistent way. Their approach helps keep the lighting changes stable and preserves the image content better than previous methods. They also created a large dataset of portraits with different lighting to train and test their model. Their experiments show that CFT works better than existing techniques and can also be used for tasks like style transfer.
image relightingdiffusion modelsfeature transportilluminationrectified flowtrajectory-level supervisionportrait datasetstyle transfer
Authors
Bohan Zhang, Huanwei Liang, Yuhan He, Hongteng Xu, Quxiao Chao, Luoqi Liu, Dixin Luo, Ting Liu
Abstract
Image relighting modifies illumination while preserving non-lighting content such as identity and geometry. Existing diffusion-based methods often suffer from unstable illumination changes or inconsistent content preservation under complex lighting, as they lack an explicit mechanism to learn feature transformations between images. We reformulate relighting as an illumination feature transport problem and introduce Consistent Feature Transport (CFT), a training principle that explicitly enforces illumination-consistent transport between source and target image distributions. Built upon rectified flow, CFT jointly models noise-to-image generation and illumination-consistent source-to-target transport through trajectory-level supervision. This dual-transport formulation encourages isolation of illumination-specific variations while preserving content-aligned features. To support complex lighting scenarios, we construct a large-scale portrait relighting dataset with diverse relighting effects. Experiments show consistent improvements over existing state-of-the-art relighting approaches and demonstrate that CFT can generalize to other editing tasks, including style transfer. Code is available at https://github.com/Dixin-Lab/CFT.