Research map / 3D vision / Face and shape
OTAvatar: One-Shot Talking Face Avatar With Controllable Tri-Plane Rendering
Zhiyuan Ma, Xiangyu Zhu, Guo-Jun Qi, Zhen Lei, Lei Zhang
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 decoupling-by-inverting rather than joint training (ablation) | OTAvatar · HDTF cross-identity reenactment | CSIM: 0.689 -> 0.694 | Improved |
| Use decoupling-by-inverting rather than joint training (ablation) | OTAvatar · HDTF cross-identity reenactment | AED: 3.009 -> 2.850 | Improved |
| Add landmark-region loss (ablation) | OTAvatar · HDTF cross-identity reenactment | AED: 3.101 -> 2.850 | Improved |
| Add ID loss (ablation) | OTAvatar · HDTF cross-identity reenactment | CSIM: 0.592 -> 0.694 | Improved |
| Add ID loss (ablation) | OTAvatar · HDTF cross-identity reenactment | AED: 3.164 -> 2.850 | Improved |
| Fine-tune the EG3D generator after controller training (ablation) | OTAvatar · HDTF cross-identity reenactment | CSIM: 0.687 -> 0.694 | Improved |
| Fine-tune the EG3D generator after controller training (ablation) | OTAvatar · HDTF cross-identity reenactment | AED: 3.145 -> 2.850 | Improved |
| Use separate latent motion codes instead of one W-space code repeated across slots (ablation) | OTAvatar · HDTF cross-identity reenactment | AED: 3.352 -> 2.850 | Improved |
| Use separate latent motion codes instead of one W-space code repeated across slots (ablation) | OTAvatar · HDTF cross-identity reenactment | CSIM: 0.719 -> 0.694 | Mixed |
| Use a learnable orthogonal codebook instead of vanilla linear weights (ablation) | OTAvatar controller · HDTF cross-identity reenactment | AED, APD, AKD, FID | Improved |
| Set identity optimization/controller training steps to 90/10 rather than other tested splits (ablation) | OTAvatar · HDTF cross-identity reenactment | AED and CSIM: 3.285 / 0.7200; 3.178 / 0.716; 2.850 / 0.694; 3.231 / 0.592 | Mixed |
Later papers that compared with this method
1 later accepted paper reports OTAvatar as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- VOODOO 3D: Volumetric Portrait Disentanglement For One-Shot 3D Head Reenactment (CVPR 2024 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- HeadNeRF — HeadNeRF: A Real-Time NeRF-Based Parametric Head Model (CVPR 2022) · better on 2 of 3 shared results
- IMAvatar — I M Avatar: Implicit Morphable Head Avatars From Videos (CVPR 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- FlashAvatar: High-fidelity Head Avatar with Efficient Gaussian Embedding (CVPR 2024 · similarity 0.88)
- Real3D-Portrait: One-shot Realistic 3D Talking Portrait Synthesis (ICLR 2024 · similarity 0.88)
- GPAvatar: Generalizable and Precise Head Avatar from Image(s) (ICLR 2024 · similarity 0.87)
- Arc2Avatar: Generating Expressive 3D Avatars from a Single Image via ID Guidance (CVPR 2025 · similarity 0.87)
- High-Fidelity 3D Head Avatars Reconstruction through Spatially-Varying Expression Conditioned Neural Radiance Field (AAAI 2024 · similarity 0.87)
- VASA-3D: Lifelike Audio-Driven Gaussian Head Avatars from a Single Image (NeurIPS 2025 · similarity 0.86)
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.
