Research map / 3D vision / Human pose estimation
GarmentNets: Category-Level Pose Estimation for Garments via Canonical Space Shape Completion
Cheng Chi, Shuran Song
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 |
|---|---|---|---|
| map to canonical space and complete there instead of directly in task space (ablation) | GarmentNets · CLOTH3D Dress | Dn: 48.32 -> 6.63 | Improved |
| add learned implicit warp field instead of physics simulation (ablation) | GarmentNets · CLOTH3D Dress | Dn: 17.11 -> 6.63 | Improved |
| add learned implicit warp field instead of physics simulation (ablation) | GarmentNets · CLOTH3D Top | Dc: 3.31 -> 1.54 | Improved |
| use winding number field instead of TSDF (ablation) | GarmentNets · CLOTH3D shape completion | Chamfer distance: 26.7% improvement over TSDF | Improved |
| use winding number field instead of TDF (ablation) | GarmentNets · CLOTH3D shape completion | Chamfer distance: 18.7% improvement over TDF | Improved |
| use winding number field instead of occupancy grid (ablation) | GarmentNets · CLOTH3D shape completion | Chamfer distance: 32.2% improvement over occupancy grid | Improved |
| use winding number field instead of TDF (ablation) | GarmentNets · CLOTH3D Top shape completion | Chamfer distance (cm): 1.22 -> 1.30 | Hurt |
| use classification instead of regression for NOCS (ablation) | GarmentNets · NOCS prediction | NOCS error: 0.16 -> 0.14 | Improved |
| use classification instead of regression for NOCS (symmetry-aware error) (ablation) | GarmentNets · NOCS prediction | NOCS error: 0.11 -> 0.06 | Improved |
| replace nearest-neighbor retrieval with GarmentNets | NN baseline · CLOTH3D Skirt | Dn: 20.55 -> 7.34 | Improved |
| replace nearest-neighbor retrieval with GarmentNets | NN baseline · CLOTH3D Dress | Dc: 2.09 -> 2.12 | No clear effect |
| raise occupancy grid resolution to 128^3 (ablation) | occupancy grid · CLOTH3D | training: 0.4% occupancy rate | Hurt |
Closest papers by meaning
- GaPT-DAR: Category-level Garments Pose Tracking via Integrated 2D Deformation and 3D Reconstruction (CVPR 2025 · similarity 0.91)
- Garment4D: Garment Reconstruction from Point Cloud Sequences (NeurIPS 2021 · similarity 0.91)
- CloSET: Modeling Clothed Humans on Continuous Surface With Explicit Template Decomposition (CVPR 2023 · similarity 0.90)
- Deep Fashion3D: A Dataset and Benchmark for 3D Garment Reconstruction from Single Images (ECCV 2020 · similarity 0.90)
- ClothPose: A Real-world Benchmark for Visual Analysis of Garment Pose via An Indirect Recording Solution (ICCV 2023 · similarity 0.90)
- DeepCloth: Neural Garment Representation for Shape and Style Editing (TPAMI 2022 · similarity 0.90)
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.
