Research map / 3D vision / Face and shape
DiFaReli: Diffusion Face Relighting
Puntawat Ponglertnapakorn, Nontawat Tritrong, Supasorn Suwajanakorn
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 Modulator network with rendered shading reference instead of concatenating reference to DDIM input (ablation) | DiFaReli · Multi-PIE ablation | LPIPS: 0.0868 -> 0.0789 | Improved |
| use spatial shading reference instead of non-spatial light vector (ablation) | DiFaReli · Multi-PIE ablation | DSSIM: 0.0885 -> 0.0670 | Improved |
| add shape, camera and identity non-spatial conditions (ablation) | DiFaReli · Multi-PIE ablation | LPIPS: 0.0909 -> 0.0789 | Improved |
| add shape and camera conditioning (ablation) | DiFaReli · Multi-PIE ablation | MSE: 0.0063 -> 0.0077 | Mixed |
| add ArcFace identity embedding conditioning (ablation) | DiFaReli · Multi-PIE ablation | LPIPS: 0.0847 -> 0.0789 | Improved |
| add background image conditioning (ablation) | DiFaReli · qualitative | background preservation | Improved |
| use DiFaReli instead of Hou et al. (CVPR'22) with same-subject target lighting | Multi-PIE · Multi-PIE (same subject) | LPIPS: 0.1622 -> 0.1370 | Improved |
| use DiFaReli with different-subject target lighting | Multi-PIE · Multi-PIE (different subject) | DSSIM: 0.1000 -> 0.0969 | Improved |
Closest papers by meaning
- DiFaReli++: Diffusion Face Relighting With Consistent Cast Shadows (TPAMI 2025 · similarity 0.99)
- SynthLight: Portrait Relighting with Diffusion Model by Learning to Re-render Synthetic Faces (CVPR 2025 · similarity 0.90)
- Lux Post Facto: Learning Portrait Performance Relighting with Conditional Video Diffusion and a Hybrid Dataset (CVPR 2025 · similarity 0.88)
- Neural Gaffer: Relighting Any Object via Diffusion (NeurIPS 2024 · similarity 0.87)
- IllumiNeRF: 3D Relighting Without Inverse Rendering (NeurIPS 2024 · similarity 0.87)
- Generative Multiview Relighting for 3D Reconstruction under Extreme Illumination Variation (CVPR 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.
