LightMover: Generative Light Movement with Color and Intensity Controls
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911550207426560 |
|---|---|
| author | Zhou, Gengze Wang, Tianyu Kim, Soo Ye Shu, Zhixin Yu, Xin Hold-Geoffroy, Yannick Chaturvedi, Sumit Wu, Qi Lin, Zhe Cohen, Scott |
| author_facet | Zhou, Gengze Wang, Tianyu Kim, Soo Ye Shu, Zhixin Yu, Xin Hold-Geoffroy, Yannick Chaturvedi, Sumit Wu, Qi Lin, Zhe Cohen, Scott |
| contents | We present LightMover, a framework for controllable light manipulation in single images that leverages video diffusion priors to produce physically plausible illumination changes without re-rendering the scene. We formulate light editing as a sequence-to-sequence prediction problem in visual token space: given an image and light-control tokens, the model adjusts light position, color, and intensity together with resulting reflections, shadows, and falloff from a single view. This unified treatment of spatial (movement) and appearance (color, intensity) controls improves both manipulation and illumination understanding. We further introduce an adaptive token-pruning mechanism that preserves spatially informative tokens while compactly encoding non-spatial attributes, reducing control sequence length by 41% while maintaining editing fidelity. To train our framework, we construct a scalable rendering pipeline that generates large numbers of image pairs across varied light positions, colors, and intensities while keeping the scene content consistent with the original image. LightMover enables precise, independent control over light position, color, and intensity, and achieves high PSNR and strong semantic consistency (DINO, CLIP) across different tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_27209 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | LightMover: Generative Light Movement with Color and Intensity Controls Zhou, Gengze Wang, Tianyu Kim, Soo Ye Shu, Zhixin Yu, Xin Hold-Geoffroy, Yannick Chaturvedi, Sumit Wu, Qi Lin, Zhe Cohen, Scott Computer Vision and Pattern Recognition Computation and Language Graphics Machine Learning I.4.0; I.2.10; I.3.3 We present LightMover, a framework for controllable light manipulation in single images that leverages video diffusion priors to produce physically plausible illumination changes without re-rendering the scene. We formulate light editing as a sequence-to-sequence prediction problem in visual token space: given an image and light-control tokens, the model adjusts light position, color, and intensity together with resulting reflections, shadows, and falloff from a single view. This unified treatment of spatial (movement) and appearance (color, intensity) controls improves both manipulation and illumination understanding. We further introduce an adaptive token-pruning mechanism that preserves spatially informative tokens while compactly encoding non-spatial attributes, reducing control sequence length by 41% while maintaining editing fidelity. To train our framework, we construct a scalable rendering pipeline that generates large numbers of image pairs across varied light positions, colors, and intensities while keeping the scene content consistent with the original image. LightMover enables precise, independent control over light position, color, and intensity, and achieves high PSNR and strong semantic consistency (DINO, CLIP) across different tasks. |
| title | LightMover: Generative Light Movement with Color and Intensity Controls |
| topic | Computer Vision and Pattern Recognition Computation and Language Graphics Machine Learning I.4.0; I.2.10; I.3.3 |
| url | https://arxiv.org/abs/2603.27209 |