Research map / 3D vision / Face and shape
Toward Realistic Single-View 3D Object Reconstruction With Unsupervised Learning From Multiple Images
Long-Nhat Ho, Anh Tuan Tran, Quynh Phung, Minh Hoai
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 |
|---|---|---|---|
| add multi-view cross-view consistency training to LeSym (ablation) | LeSym · BFM | MAD: 16.51 -> 15.73 | Improved |
| add multi-view training to LeSym (ablation) | LeSym · BFM | SIDE (x10^-2): 0.793 -> 0.728 | Improved |
| add albedo loss to LeSym (ablation) | LeSym · BFM | MAD: 16.51 -> 16.35 | Improved |
| add albedo loss to LeSym (ablation) | LeSym · BFM | SIDE (x10^-2): 0.793 -> 0.899 | Hurt |
| add albedo loss on top of multi-view (ablation) | LeMul · BFM | MAD: 15.73 -> 15.49 | Improved |
| add albedo loss on top of multi-view (ablation) | LeMul · BFM | SIDE (x10^-2): 0.728 -> 0.834 | Hurt |
| replace LeSym with LeMul | unsupervised 3D face reconstruction · BFM | MAD: 16.51 -> 15.49 | Improved |
| replace LeSym with LeMul | unsupervised 3D face reconstruction · BFM face region | SIDE: 0.00564 -> 0.00534 | Improved |
| replace LeSym with LeMul | unsupervised 3D face reconstruction · User survey CASIA | selection rate (%): 20.86 -> 79.14 | Improved |
| replace LeSym with LeMul | unsupervised 3D face reconstruction · User survey CatFaces | selection rate (%): 47.03 -> 52.97 | No clear effect |
Closest papers by meaning
- Unsupervised Learning of Probably Symmetric Deformable 3D Objects From Images in the Wild (CVPR 2020 · similarity 0.90)
- Front2Back: Single View 3D Shape Reconstruction via Front to Back Prediction (CVPR 2020 · similarity 0.89)
- Learning View Priors for Single-View 3D Reconstruction (CVPR 2019 · similarity 0.89)
- Single-View 3D Object Reconstruction From Shape Priors in Memory (CVPR 2021 · similarity 0.89)
- Unsupervised Learning of 3D Structure from Images (NeurIPS 2016 · similarity 0.88)
- Learning Partonomic 3D Reconstruction from Image Collections (CVPR 2025 · similarity 0.87)
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.
