⭐️ Our team's works: [MotionClone] [BroadWay]
This repository is the official implementation of Light-A-Video. It is a training-free framework that enables zero-shot illumination control of any given video sequences or foreground sequences.
Click for the full abstract of Light-A-Video
Recent advancements in image relighting models, driven by large-scale datasets and pre-trained diffusion models, have enabled the imposition of consistent lighting. However, video relighting still lags, primarily due to the excessive training costs and the scarcity of diverse, high-quality video relighting datasets. A simple application of image relighting models on a frame-by-frame basis leads to several issues: lighting source inconsistency and relighted appearance inconsistency, resulting in flickers in the generated videos. In this work, we propose Light-A-Video, a training-free approach to achieve temporally smooth video relighting. Adapted from image relighting models, Light-A-Video introduces two key techniques to enhance lighting consistency. First, we design a Consistent Light Attention (CLA) module, which enhances cross-frame interactions within the self-attention layers to stabilize the generation of the background lighting source. Second, leveraging the physical principle of light transport independence, we apply linear blending between the source video’s appearance and the relighted appearance, using a Progressive Light Fusion (PLF) strategy to ensure smooth temporal transitions in illumination. Experiments show that Light-A-Video improves the temporal consistency of relighted video while maintaining the image quality, ensuring coherent lighting transitions across frames.
Light-A-Video: Training-free Video Relighting via Progressive Light Fusion
Yujie Zhou*,
Jiazi Bu*,
Pengyang Ling*,
Pan Zhang†,
Tong Wu,
Qidong Huang,
Jinsong Li,
Xiaoyi Dong,
Yuhang Zang,
Yuhang Cao,
Anyi Rao,
Jiaqi Wang,
Li Niu†
(*Equal Contribution)(†Corresponding Author)
[2025/2/11] Code is available now!
[2025/2/10] The paper and project page are released!
-
Release a gradio demo.
-
Release Light-A-Video code with CogVideoX-2B pipeline.
We show more results in the Project Page.
..., red and blue neon light |
..., sunset over sea |
 |
 |
..., sunlight through the blinds |
..., in the forest, magic golden lit |
 |
 |
Light-A-Video leverages the the capabilities of image relighting models and VDM motion priors to achieve temporally consistent video relighting. By integrating the Consistent Light Attention to stabilize lighting source generation and employ the Progressive Light Fusion strategy for smooth appearance transitions.
git clone https://github.com/bcmi/Light-A-Video.git
cd Light-A-Video
conda create -n lav python=3.10
conda activate lav
pip install -r requirements.txt- IC-Light: Huggingface
- SD RealisticVision: Huggingface
- Animatediff Motion-Adapter-V-1.5.3: Huggingface
Model downloading is automatic.
# relight
python lav_relight.py --config "configs/relight/car.yaml"A script based on SAM2 is provided to extract foreground sequences from videos.
# extract foreground sequence
python sam2.py --video_name car --x 255 --y 255
# inpaint and relight
python lav_paint.py --config "configs/relight_inpaint/car.yaml"If you find our work helpful for your research, please consider giving a star ⭐ and citation 📝
@article{zhou2025light,
title={Light-A-Video: Training-free Video Relighting via Progressive Light Fusion},
author={Zhou, Yujie and Bu, Jiazi and Ling, Pengyang and Zhang, Pan and Wu, Tong and Huang, Qidong and Li, Jinsong and Dong, Xiaoyi and Zang, Yuhang and Cao, Yuhang and others},
journal={arXiv preprint arXiv:2502.08590},
year={2025}
}This is official code of Light-A-Video. All the copyrights of the demo images and audio are from community users. Feel free to contact us if you would like remove them.
The code is built upon the below repositories, we thank all the contributors for open-sourcing.