Research map / 3D vision / NeRF and radiance fields
One-Shot High-Fidelity Talking-Head Synthesis With Deformable Neural Radiance Field
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 HiDe-NeRF instead of prior methods | talking-head cross-identity reenactment · VoxCeleb1, VoxCeleb2, and TalkingHead-1KH | CSIM: 0.833 -> 0.876; 0.711 -> 0.787; 0.781 -> 0.828 | Helped |
| use HiDe-NeRF instead of prior methods | talking-head cross-identity reenactment · VoxCeleb1, VoxCeleb2, and TalkingHead-1KH | AUCON: 0.917; 0.889; 0.901 | Helped |
| use HiDe-NeRF instead of prior methods | talking-head cross-identity reenactment · VoxCeleb1, VoxCeleb2, and TalkingHead-1KH | FID: 57; 61; 52 | Helped |
| use HiDe-NeRF instead of OSFV-S | cross-identity free-view generation · VoxCeleb1 | CSIM: 0.727 -> 0.829 | Helped |
| use multi-scale instead of single-scale tri-plane representations (ablation) | HiDe-NeRF · VoxCeleb1 ablation study | CSIM: 0.814 -> 0.876; 0.062 | Helped |
| use multi-scale instead of single-scale tri-plane representations (ablation) | HiDe-NeRF · VoxCeleb1 ablation study | AUCON: 0.852 -> 0.917; 0.065 | Helped |
| use multi-scale instead of single-scale tri-plane representations (ablation) | HiDe-NeRF · VoxCeleb1 ablation study | PRMSE: 2.62 -> 2.62 | No clear effect |
| add camera-to-world transformation (ablation) | HiDe-NeRF · VoxCeleb1 ablation study | CSIM: 0.760 -> 0.876 | Helped |
| add camera-to-world transformation (ablation) | HiDe-NeRF · VoxCeleb1 ablation study | PRMSE: 3.58 -> 2.62 | Helped |
| use pose-agnostic SECC instead of pose-coupled SECC (ablation) | HiDe-NeRF · VoxCeleb1 ablation study | CSIM: 0.802 -> 0.876; 0.074 | Helped |
| use pose-agnostic SECC instead of pose-coupled SECC (ablation) | HiDe-NeRF · VoxCeleb1 ablation study | AUCON: 0.879 -> 0.917; 0.038 | Helped |
| use pose-agnostic SECC instead of pose-coupled SECC (ablation) | HiDe-NeRF · VoxCeleb1 ablation study | AVD: 0.018 -> 0.012; 0.006 | Helped |
Closest papers by meaning
- AE-NeRF: Audio Enhanced Neural Radiance Field for Few Shot Talking Head Synthesis (AAAI 2024 · similarity 0.90)
- High-Fidelity 3D Head Avatars Reconstruction through Spatially-Varying Expression Conditioned Neural Radiance Field (AAAI 2024 · similarity 0.88)
- FaceTalk: Audio-Driven Motion Diffusion for Neural Parametric Head Models (CVPR 2024 · similarity 0.88)
- SyncTalk: The Devil is in the Synchronization for Talking Head Synthesis (CVPR 2024 · similarity 0.88)
- MimicTalk: Mimicking a personalized and expressive 3D talking face in minutes (NeurIPS 2024 · similarity 0.88)
- Hierarchically Controlled Deformable 3D Gaussians for Talking Head Synthesis (AAAI 2025 · similarity 0.88)
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.