Research map / Video / Video generation
Personalized Video Relighting With an At-Home Light Stage
Jun Myeong Choi, Max Christman, Roni Sengupta
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 LCFN de-lighting of skip-connected features (ablation) | relighting U-Net · LSYD | LPIPS: 0.0964 -> 0.0904 | Improved |
| add source monitor lighting prediction from intermediate encoder features (ablation) | relighting U-Net · LSYD | LPIPS: 0.0964 -> 0.0915 | Improved |
| add both LCFN and source lighting prediction (ablation) | relighting U-Net · LSYD | LPIPS: 0.0964 -> 0.0853 | Improved |
| add skip-connected feature smoothing (exponential smoothing of de-lit LCFN features) (ablation) | relighting U-Net · LSYD temporal consistency | adjacent-frame RMSE: 6.01 -> 5.68 | Improved |
| add source monitor lighting averaging over previous frames (ablation) | relighting U-Net · LSYD temporal consistency | adjacent-frame RMSE: 6.01 -> 5.73 | Improved |
| add source-light averaging to Sengupta et al. (ablation) | Sengupta et al. · LSYD temporal consistency | adjacent-frame RMSE: 6.37 -> 5.76 | Improved |
| use the proposed loss handling pose misalignment when training Sun et al. (ablation) | Sun et al. · LSYD | LPIPS: 0.1712 -> 0.1029 | Improved |
| use predicted source lighting instead of known source lighting (ablation) | proposed relighting network · LSYD | LPIPS: 0.0832 -> 0.0839 | No clear effect |
| use segmentation versus facial keypoints for pairing source and target images (ablation) | training data construction · LSYD | pair quality | Improved |
Closest papers by meaning
- Lux Post Facto: Learning Portrait Performance Relighting with Conditional Video Diffusion and a Hybrid Dataset (CVPR 2025 · similarity 0.90)
- Comprehensive Relighting: Generalizable and Consistent Monocular Human Relighting and Harmonization (CVPR 2025 · similarity 0.90)
- Real-time 3D-aware Portrait Video Relighting (CVPR 2024 · similarity 0.89)
- DreamLight: Towards Harmonious and Consistent Image Relighting (NeurIPS 2025 · similarity 0.88)
- SynthLight: Portrait Relighting with Diffusion Model by Learning to Re-render Synthetic Faces (CVPR 2025 · similarity 0.87)
- UniLumos: Fast and Unified Image and Video Relighting with Physics-Plausible Feedback (NeurIPS 2025 · similarity 0.87)
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.
