Research map / 3D vision / Face and shape
Fast Bi-layer Neural Synthesis of One-Shot Realistic Head Avatars
Egor Zakharov, Aleksei Ivakhnenko, Aliaksandra Shysheya, Victor Lempitsky
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 AdaSPADE and adaptive skip-connections (ablation) | bi-layer avatar model · VoxCeleb2 | CSIM: 0.547 -> 0.595 | Improved |
| add AdaSPADE and adaptive skip-connections (ablation) | bi-layer avatar model · VoxCeleb2 | LPIPS: 0.377 -> 0.370 | Improved |
| add AdaSPADE and adaptive skip-connections (ablation) | bi-layer avatar model · VoxCeleb2 | NME: 0.447 -> 0.441 | Improved |
| add updater network (ablation) | bi-layer avatar model · VoxCeleb2 | CSIM: 0.595 -> 0.653 | Improved |
| add updater network (ablation) | bi-layer avatar model · VoxCeleb2 | LPIPS: 0.370 -> 0.358 | Improved |
| add updater network (ablation) | bi-layer avatar model · VoxCeleb2 | NME: 0.441 -> 0.433 | Improved |
| add updater network (ablation) | bi-layer avatar model · VoxCeleb2-HQ | CSIM: 0.432 -> 0.649 | Improved |
| add updater network (ablation) | bi-layer avatar model · VoxCeleb2-HQ | LPIPS: 0.313 -> 0.298 | Improved |
| add updater network (ablation) | bi-layer avatar model · VoxCeleb2-HQ | NME: 0.476 -> 0.456 | Improved |
| use bi-layer model instead of FOMM at medium size | medium models · VoxCeleb2 | GMACs: 13.9 -> 4.32 | Improved |
| use bi-layer model instead of FOMM at medium size | medium models · VoxCeleb2 | LPIPS: 0.311 -> 0.358 | Hurt |
| use bi-layer model instead of Few-shot Vid2Vid at medium size | medium models · VoxCeleb2 | NME: 0.461 -> 0.433 | Improved |
Closest papers by meaning
- Efficient 3D Implicit Head Avatar with Mesh-anchored Hash Table Blendshapes (CVPR 2024 · similarity 0.89)
- Synthetic Prior for Few-Shot Drivable Head Avatar Inversion (CVPR 2025 · similarity 0.88)
- Generalizable and Animatable Gaussian Head Avatar (NeurIPS 2024 · similarity 0.88)
- MobilePortrait: Real-Time One-Shot Neural Head Avatars on Mobile Devices (CVPR 2025 · similarity 0.87)
- Artist-Friendly Relightable and Animatable Neural Heads (CVPR 2024 · similarity 0.87)
- Gaussian Eigen Models for Human Heads (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.
