Research map / 3D vision / Face and shape
PointAvatar: Deformable Point-Based Head Avatars From Videos
Yufeng Zheng, Wang Yifan, Gordon Wetzstein, Michael J. Black, Otmar Hilliges
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 deformable point-based avatars instead of implicit avatar representations | PointAvatar · Lab-capture sequences | LPIPS: 0.178; 0.126 -> 0.094 | Improved |
| Use deformable point-based avatars instead of implicit avatar representations | PointAvatar · Casual videos | LPIPS: 0.261; 0.187 -> 0.145 | Improved |
| Use deformable point-based avatars instead of implicit avatar representations | PointAvatar · Lab-capture sequences | PSNR: 22.4; 22.7 -> 26.6 | Improved |
| Use deformable point-based avatars instead of implicit avatar representations | PointAvatar · Casual videos | PSNR: 18.7; 19.6 -> 22.3 | Improved |
| Add albedo-shading disentanglement (ablation) | PointAvatar · Facial geometry | geometry detail | Improved |
| Estimate canonical point normals from an SDF rather than neighboring-point planes (ablation) | PointAvatar · Detailed regions including hair, eyes, and mouth | normal smoothness | Improved |
| Transform normals with the deformation Jacobian rather than only the rotation matrix (ablation) | PointAvatar · Deformed facial geometry, including the nasal line when smiling | normal accuracy | Improved |
| Add VGG feature loss to per-pixel L1 loss (ablation) | PointAvatar · Avatar rendering | photo-realism | Improved |
| Add a freely learned canonical offset instead of forcing points into predefined FLAME canonical space (ablation) | PointAvatar · Canonical geometry and novel head poses | generalization | Improved |
| Use two-stage deformation rather than a single-stage mapping (ablation) | PointAvatar · Avatar geometry optimization | geometry accuracy | Improved |
| Use a point-based representation rather than a mesh-based representation | PointAvatar · Eyeglasses and thin hair structures | geometry detail | Improved |
Later papers that compared with this method
15 later accepted papers report PointAvatar as a baseline in their results tables; 15 of them report a better number on most of the results they share with it.
- DipGuava: Disentangling Personalized Gaussian Features for 3D Head Avatars from Monocular Video (AAAI 2026 · better on 4 of 4 shared results)
- ProgressiveAvatars: Progressive Animatable 3D Gaussian Avatars (CVPR 2026 · better on 2 of 2 shared results)
- MeGA: Hybrid Mesh-Gaussian Head Avatar for High-Fidelity Rendering and Head Editing (CVPR 2025 · better on 6 of 6 shared results)
- SVG-Head: Hybrid Surface-Volumetric Gaussians for High-Fidelity Head Reconstruction and Real-Time Editing (ICCV 2025 · better on 4 of 4 shared results)
- SurFhead: Affine Rig Blending for Geometrically Accurate 2D Gaussian Surfel Head Avatars (ICLR 2025 · better on 2 of 2 shared results)
- Gaussian Head & Shoulders: High Fidelity Neural Upper Body Avatars with Anchor Gaussian Guided Texture Warping (ICLR 2025 · better on 2 of 2 shared results)
- HRAvatar: High-Quality and Relightable Gaussian Head Avatar (CVPR 2025 · better on 1 of 1 shared result)
- GaussianAvatars: Photorealistic Head Avatars with Rigged 3D Gaussians (CVPR 2024 · better on 6 of 6 shared results)
7 more not listed.
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- IMavatar — I M Avatar: Implicit Morphable Head Avatars From Videos (CVPR 2022) · better on 4 of 4 shared results
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
- HRAvatar: High-Quality and Relightable Gaussian Head Avatar (CVPR 2025 · similarity 0.90)
- GPAvatar: High-fidelity Head Avatars by Learning Efficient Gaussian Projections (CVPR 2025 · similarity 0.89)
- GPAvatar: Generalizable and Precise Head Avatar from Image(s) (ICLR 2024 · similarity 0.89)
- FlashAvatar: High-fidelity Head Avatar with Efficient Gaussian Embedding (CVPR 2024 · similarity 0.88)
- HERA: Hybrid Explicit Representation for Ultra-Realistic Head Avatars (CVPR 2025 · similarity 0.88)
- RGBAvatar: Reduced Gaussian Blendshapes for Online Modeling of Head Avatars (CVPR 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.
