Research map / 3D vision / 3D generation
LuminAIRe: Illumination-Aware Conditional Image Repainting for Lighting-Realistic Generation
Jiajun Tang, Haofeng Zhong, Shuchen Weng, Boxin Shi
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 |
|---|---|---|---|
| add hierarchical labeling enhancement (coarsened masks in training) (ablation) | LuminAIRe · CAR-LuminAIRe | FID: 5.83 -> 4.62 | Improved |
| add hierarchical labeling enhancement (ablation) | LuminAIRe · CAR-LuminAIRe | R-precision: 63.27% -> 74.13% | Improved |
| add hierarchical labeling enhancement (ablation) | LuminAIRe · CAR-LuminAIRe | M-score: 13.97 -> 13.68 | Improved |
| add illumination attention module (ablation) | LuminAIRe · CAR-LuminAIRe | FID: 6.31 -> 5.83 | Mixed |
| add illumination attention module (ablation) | LuminAIRe · CAR-LuminAIRe | SSIM: 0.7214 -> 0.7163 | No clear effect |
| add illumination injection for the foreground (ablation) | LuminAIRe · CAR-LuminAIRe | FID: 8.00 -> 6.31 | Improved |
| add illumination injection for the foreground (ablation) | LuminAIRe · CAR-LuminAIRe | M-score: 15.83 -> 13.95 | Improved |
| add illumination injection for the foreground (ablation) | LuminAIRe · CAR-LuminAIRe | SSIM: 0.7054 -> 0.7214 | Improved |
| add lighting-realistic background generation (UniCoRN to Ours-HAI) (ablation) | UniCoRN · CAR-LuminAIRe | FID: 11.55 -> 8.00 | Improved |
| add lighting-realistic background generation (UniCoRN to Ours-HAI) (ablation) | UniCoRN · CAR-LuminAIRe | M-score: 29.72 -> 15.83 | Improved |
| LuminAIRe vs UniCoRN | UniCoRN · CAR-LuminAIRe | FID: 11.55 -> 4.62 | Improved |
| LuminAIRe vs Pavllo et al. | Pavllo et al. · CAR-LuminAIRe | User study realistic: 43.00% -> 48.12% | Improved |
Closest papers by meaning
- UniLumos: Fast and Unified Image and Video Relighting with Physics-Plausible Feedback (NeurIPS 2025 · similarity 0.87)
- UniRelight: Learning Joint Decomposition and Synthesis for Video Relighting (NeurIPS 2025 · similarity 0.86)
- LumiNet: Latent Intrinsics Meets Diffusion Models for Indoor Scene Relighting (CVPR 2025 · similarity 0.86)
- LightIt: Illumination Modeling and Control for Diffusion Models (CVPR 2024 · similarity 0.86)
- Luminance-GS: Adapting 3D Gaussian Splatting to Challenging Lighting Conditions with View-Adaptive Curve Adjustment (CVPR 2025 · similarity 0.86)
- Luminance-Aware Statistical Quantization: Unsupervised Hierarchical Learning for Illumination Enhancement (NeurIPS 2025 · similarity 0.86)
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.
