Research map / Video / Video generation
ReliTalk: Relightable Talking Portrait Generation from a Single Video
Haonan Qiu, Zhaoxi Chen, Yuming Jiang, Hang Zhou, Xiangyu Fan, Lei Yang, Wayne Wu, Ziwei Liu
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 mesh-aware guidance to audio features (ablation) | ReliTalk audio-to-expression · Audio-to-expression prediction | PSNR: 29.6147 -> 34.1099 | Improved |
| add phoneme audio features to mesh guidance (ablation) | ReliTalk audio-to-expression · Audio-to-expression prediction | PSNR: 33.8643 -> 34.1099 | Improved |
| add mesh-aware guidance when training video is short (ablation) | ReliTalk · Short training video (2450 frames) | PSNR: 29.5445 -> 34.8186 | Improved |
| add identity-consistent supervision with simulated lighting (ablation) | ReliTalk normal estimation · Normal map quality | PSNR_grad: 21.9828 -> 23.5441 | Improved |
| add identity-consistent supervision with simulated lighting (ablation) | ReliTalk normal estimation · Normal map quality | SSIM: 0.9060 -> 0.9203 | Improved |
| add identity-consistent supervision with simulated lighting (ablation) | ReliTalk normal estimation · Normal map quality | PSNR: 21.7141 -> 21.5835 | No clear effect |
| add initial normal from audio-to-geometry (ablation) | ReliTalk reflectance decomposition · Relighting visual results | convergence | Improved |
| add normal residual network (ablation) | ReliTalk reflectance decomposition · Relighting visual results | relighting quality | Improved |
| add parsimony prior on albedo (ablation) | ReliTalk reflectance decomposition · Relighting visual results | relighting vividness | Improved |
| add smoothness constraints on decomposed maps (ablation) | ReliTalk reflectance decomposition · Relighting visual results | relighting vividness | Improved |
| add specular weight network (ablation) | ReliTalk reflectance decomposition · Relighting visual results | sharpness | Improved |
| add specular map via Blinn-Phong (ablation) | ReliTalk reflectance decomposition · Relighting visual results | photorealism | Improved |
Closest papers by meaning
- Real-time 3D-aware Portrait Video Relighting (CVPR 2024 · similarity 0.90)
- Real3D-Portrait: One-shot Realistic 3D Talking Portrait Synthesis (ICLR 2024 · similarity 0.89)
- DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking head Video Generation (ICLR 2025 · similarity 0.88)
- GoHD: Gaze-oriented and Highly Disentangled Portrait Animation with Rhythmic Poses and Realistic Expressions (AAAI 2025 · similarity 0.88)
- Lux Post Facto: Learning Portrait Performance Relighting with Conditional Video Diffusion and a Hybrid Dataset (CVPR 2025 · similarity 0.87)
- High-Fidelity Relightable Monocular Portrait Animation with Lighting-Controllable Video Diffusion Model (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.
