Research map / 3D vision / Face and shape
Learning Anchored Unsigned Distance Functions With Gradient Direction Alignment for Single-View Garment Reconstruction
Fang Zhao, Wenhao Wang, Shengcai Liao, Ling Shao
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 the gradient direction alignment loss L_GDA (ablation) | AnchorUDF · MGN | Chamfer: 0.696 -> 0.635 | Improved |
| add the gradient direction alignment loss L_GDA (ablation) | AnchorUDF · MGN | P2S: 0.899 -> 0.762 | Improved |
| add the gradient direction alignment loss L_GDA (ablation) | AnchorUDF · Deep Fashion3D | Chamfer: 0.712 -> 0.621 | Improved |
| add the gradient direction alignment loss L_GDA (ablation) | AnchorUDF · Deep Fashion3D | P2S: 0.932 -> 0.839 | Improved |
| use anchor-point position features instead of depth value (ablation) | AnchorUDF · MGN | Chamfer: 1.063 -> 0.696 | Improved |
| use anchor-point position features instead of 3D coordinates (ablation) | AnchorUDF · MGN | P2S: 1.060 -> 0.899 | Improved |
| use anchor-point position features instead of depth value (ablation) | AnchorUDF · Deep Fashion3D | P2S: 2.440 -> 0.932 | Improved |
| use anchor-point position features instead of 3D coordinates (ablation) | AnchorUDF · Deep Fashion3D | Chamfer: 1.062 -> 0.712 | Improved |
| increase anchor points from 100 to 600 (ablation) | AnchorUDF · MGN | Chamfer: 0.731 -> 0.696 | Improved |
| increase anchor points from 600 to 900 (ablation) | AnchorUDF · MGN | Chamfer: 0.696 -> 0.758 | Hurt |
| increase anchor points from 600 to 900 (ablation) | AnchorUDF · MGN | P2S: 0.899 -> 1.014 | Hurt |
| add the HD module with 1024x1024 input | AnchorUDF · Deep Fashion3D | P2S: 0.839 -> 0.737 | Improved |
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.
- BCNet — BCNet: Learning Body and Cloth Shape from A Single Image (ECCV 2020) · better on 2 of 2 shared results
- Pixel2Mesh — Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images (ECCV 2018) · better on 2 of 2 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
- Registering Explicit to Implicit: Towards High-Fidelity Garment Mesh Reconstruction From Single Images (CVPR 2022 · similarity 0.89)
- Deep Fashion3D: A Dataset and Benchmark for 3D Garment Reconstruction from Single Images (ECCV 2020 · similarity 0.88)
- Garment4D: Garment Reconstruction from Point Cloud Sequences (NeurIPS 2021 · similarity 0.88)
- VS: Reconstructing Clothed 3D Human from Single Image via Vertex Shift (CVPR 2024 · similarity 0.87)
- SIFU: Side-view Conditioned Implicit Function for Real-world Usable Clothed Human Reconstruction (CVPR 2024 · similarity 0.86)
- DressWild: Feed-Forward Pose-Agnostic Garment Sewing Pattern Generation from In-the-Wild Images (NeurIPS 2026 · 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.
