Research map / Image generation / Editing and style transfer
TokenLight: Precise Lighting Control in Images using Attribute Tokens
Sumit Chaturvedi, Yannick Hold-Geoffroy, Mengwei Ren, Jingyuan Liu, He Zhang, Yiqun Mei, Julie Dorsey, Zhixin Shu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use tokenized point-light control instead of environment-map relighting (Neural Gaffer) | TokenLight · Synthetic PointGT | PSNR: 16.7595 -> 21.9772 | Improved |
| use tokenized control instead of DiffusionRenderer | TokenLight · Synthetic PanoGT | PSNR: 13.5106 -> 21.2416 | Improved |
| use tokenized control instead of Neural Gaffer | TokenLight · Synthetic PointGT | LPIPS: 0.0415 -> 0.0210 | Improved |
| end-to-end tokens instead of 2.5D reconstruction pipeline (ScribbleLight) | TokenLight · VisibleFixture-60 | PSNR: 14.6356 -> 20.0772 | Improved |
| end-to-end tokens instead of ScribbleLight | TokenLight · VisibleFixture-60 | LPIPS: 0.6114 -> 0.2784 | Improved |
| attribute tokens instead of Neural Gaffer for light position precision | TokenLight · Precision analysis | B/A sensitivity: 1.111 -> 1.877 | Improved |
| attribute tokens instead of Neural Gaffer for light position | TokenLight · Precision analysis | A accuracy error: 0.160 -> 0.049 | Improved |
| camera-tied token lighting instead of GenLit view-agnostic parameterization | TokenLight · User study | preference: 77.5% | Improved |
| end-to-end tokens instead of Careaga 2.5D pipeline | TokenLight · User study | preference: 89.2% | Improved |
| increase sampling steps 1 to 50 | TokenLight · Synthetic set | LPIPS and SSIM | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Neural Gaffer — Neural Gaffer: Relighting Any Object via Diffusion (NeurIPS 2024) · better on 8 of 8 shared results
- ScribbleLight — ScribbleLight: Single Image Indoor Relighting with Scribbles (CVPR 2025) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- UniLumos: Fast and Unified Image and Video Relighting with Physics-Plausible Feedback (NeurIPS 2025 · similarity 0.89)
- DreamLight: Towards Harmonious and Consistent Image Relighting (NeurIPS 2025 · similarity 0.89)
- UniRelight: Learning Joint Decomposition and Synthesis for Video Relighting (NeurIPS 2025 · similarity 0.89)
- LumiNet: Latent Intrinsics Meets Diffusion Models for Indoor Scene Relighting (CVPR 2025 · similarity 0.89)
- SynthLight: Portrait Relighting with Diffusion Model by Learning to Re-render Synthetic Faces (CVPR 2025 · similarity 0.88)
- LightIt: Illumination Modeling and Control for Diffusion Models (CVPR 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
