Research map / 3D vision / Face and shape
Detailed Facial Geometry Recovery from Multi-View Images by Learning an Implicit Function
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 mesh-attached UVD cost space and implicit-function regularization rather than generalized camera-frustum MVS | Facial multi-view reconstruction · FaceScape test set | CD-mean: 0.305 -> 0.175 | Helped |
| Use implicit-function prediction instead of 3D CNN regularization (ablation) | FaceScape reconstruction · FaceScape test set | CD-mean (mm): 0.230 -> 0.195 | Helped |
| Add a post-regularizer to the implicit-function framework (ablation) | Base+IF · FaceScape test set | CD-mean (mm): 0.195 -> 0.171 | Helped |
| Use distance-to-surface labels instead of inside/outside labels (ablation) | Implicit-function reconstruction with post-regularization · FaceScape test set | CD-mean (mm): 0.181 -> 0.171 | Helped |
| Add mesoscopic prediction (ablation) | Base+IF+Reg · FaceScape test set | CD-mean (mm): 0.171 -> 0.175 | Hurt |
| Add mesoscopic prediction (ablation) | Base+IF+Reg · FaceScape test set | visual mesoscopic detail | Helped |
| Use mesh-attached UVD cost space rather than camera-frustum cost space (ablation) | Facial MVS · FaceScape test set | CD-mean: 0.305 -> 0.230 | Helped |
| Reduce input views from 10 to 2 (ablation) | Proposed method · FaceScape test set | CD-mean (mm): 0.167 -> 5.264 | Hurt |
Closest papers by meaning
- Deep Facial Non-Rigid Multi-View Stereo (CVPR 2020 · similarity 0.90)
- MVF-Net: Multi-View 3D Face Morphable Model Regression (CVPR 2019 · similarity 0.90)
- Learning Detailed Face Reconstruction From a Single Image (CVPR 2017 · similarity 0.89)
- A Hierarchical Representation Network for Accurate and Detailed Face Reconstruction From In-the-Wild Images (CVPR 2023 · similarity 0.88)
- Lightweight Photometric Stereo for Facial Details Recovery (CVPR 2020 · similarity 0.88)
- Anatomically Constrained Implicit Face Models (CVPR 2024 · similarity 0.85)
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.