Research map / 3D vision / Face and shape
WarpHE4D: Dense 4D Head Map toward Full Head Reconstruction
Jongseob Yun, Yong-Hoon Kwon, Min-Gyu Park, Ju-Mi Kang, Min-Ho Lee, Inho Chang, Ju Hong Yoon, Kuk-Jin Yoon
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 |
|---|---|---|---|
| replace DINOv2 with VGG features only (no self-supervised features) (ablation) | WarpHE4D · Multiface, face region | UVAUC: 0.7864 -> 0.5252 | Hurt |
| use Sapiens human-centric foundation model instead of DINOv2 (ablation) | WarpHE4D · Multiface, face region | UVAUC: 0.7864 -> 0.1282 | Hurt |
| add keypoint loss on UV predictions (ablation) | WarpHE4D · Multiface, face region | UVAUC: 0.7695 -> 0.7864 | Improved |
| add keypoint loss on UV predictions (ablation) | WarpHE4D · Multiface, face region | 3DAUC: 0.8846 -> 0.9115 | Improved |
| add keypoint loss on UV predictions (ablation) | WarpHE4D · Multiface, non-facial region | MP3DE: 0.0204 -> 0.0105 | Improved |
| use DINOv2-based head map regression instead of 3DMM-based DECA in non-facial regions | head reconstruction · Multiface, non-facial | UVAUC: 0.6022 -> 0.6858 | Improved |
| DINOv2 head map vs DECA in face region under Procrustes-aligned baselines | face region · Multiface, face region | UVAUC: 0.8073 -> 0.7864 | Mixed |
| WarpHE4D vs baselines across head poses on FaceVerse | head reconstruction · FaceVerse | UVAUC: min margin 0.09, max margin 0.23 | Improved |
| restrict training views to azimuth within [-90, 90] degrees | training set · overall | performance | Improved |
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.
- EMOCA — EMOCA: Emotion Driven Monocular Face Capture and Animation (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
- GeoDiff4D: Geometry-Aware Diffusion for 4D Head Avatar Reconstruction (CVPR 2026 · similarity 0.88)
- FFAvatar: Feed-Forward 4D Head Avatar Reconstruction from Arbitrary Images (ECCV 2026 · similarity 0.88)
- Topo4D++: Realistic Physically Based 4D Head Capture With Topology-Preserving Gaussian Splatting and Expression Priors (TPAMI 2025 · similarity 0.87)
- RGBAvatar: Reduced Gaussian Blendshapes for Online Modeling of Head Avatars (CVPR 2025 · similarity 0.86)
- 3D Gaussian Head Avatars with Expressive Dynamic Appearances by Compact Tensorial Representations (CVPR 2025 · similarity 0.86)
- Synthetic Prior for Few-Shot Drivable Head Avatar Inversion (CVPR 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.
