Research map / 3D vision / Face and shape
DeePSD: Automatic Deep Skinning and Pose Space Deformation for 3D Garment Animation
Hugo Bertiche, Meysam Madadi, Emilio Tylson, Sergio Escalera
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 |
|---|---|---|---|
| predict blend shape matrices (PSD) instead of vertex deformations, with global descriptor (ablation) | Baseline · CLOTH3D subsample (50k train/5k test frames) | Euclidean error (mm): 29.98 -> 25.13 | Improved |
| add global outfit descriptor to the deformation baseline (ablation) | Baseline · CLOTH3D subsample (50k train/5k test frames) | Euclidean error (mm): 29.98 -> 28.04 | Improved |
| lower global descriptor dimensionality (F=128) (ablation) | +Global · CLOTH3D subsample (50k train/5k test frames) | Euclidean error (mm): 28.04 -> 28.59 | Hurt |
| use graph convolutions instead of fully-connected layers after the encoder (ablation) | +Global · CLOTH3D subsample (50k train/5k test frames) | Euclidean error (mm): 28.04 -> 28.76 | Hurt |
| replace single FC layer by MLP for global descriptor (ablation) | +Global · CLOTH3D subsample (50k train/5k test frames) | Euclidean error (mm): 28.04 -> 28.43 | No clear effect |
| add pose embedding (ablation) | DeePSD · CLOTH3D subsample (50k train/5k test frames) | Euclidean error (mm): 30.93 -> 25.13 | Improved |
| add SMPL shape and gender metadata (ablation) | DeePSD · CLOTH3D subsample (50k train/5k test frames), metadata conditioning | Euclidean error (mm): 25.13 -> 25.15 | No clear effect |
| add garment fabric metadata (ablation) | DeePSD · CLOTH3D subsample (50k train/5k test frames), metadata conditioning | Euclidean error (mm): 25.13 -> 24.76 | Improved |
| add fabric and tightness metadata (ablation) | DeePSD · CLOTH3D subsample (50k train/5k test frames), metadata conditioning | Euclidean error (mm): 25.13 -> 24.66 | Improved |
| add all metadata including SMPL body info (ablation) | DeePSD · CLOTH3D subsample (50k train/5k test frames), metadata conditioning | Euclidean error (mm): 25.13 -> 25.01 | No clear effect |
| add unsupervised physical consistency loss (ablation) | DeePSD · CLOTH3D subsample (50k train/5k test frames), physical consistency | Collision (% vertices): 11.59% -> 1.29% | Improved |
| add unsupervised physical consistency loss (ablation) | DeePSD · CLOTH3D subsample (50k train/5k test frames), physical consistency | Euclidean error (mm): 24.66 -> 33.75 | Hurt |
Later papers that compared with this method
3 later accepted papers report DeePSD as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- PhysDrape: Learning Explicit Forces and Collision Constraints for Physically Realistic Garment Draping (ECCV 2026 · better on 3 of 3 shared results)
- Towards Multi-Layered 3D Garments Animation (ICCV 2023 · better on 8 of 9 shared results)
- DrapeNet: Garment Generation and Self-Supervised Draping (CVPR 2023 · better on 3 of 5 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
- DeepCloth: Neural Garment Representation for Shape and Style Editing (TPAMI 2022 · similarity 0.93)
- Learning 3D Garment Animation from Trajectories of A Piece of Cloth (NeurIPS 2024 · similarity 0.90)
- Self-Supervised Collision Handling via Generative 3D Garment Models for Virtual Try-On (CVPR 2021 · similarity 0.90)
- DeepWrinkles: Accurate and Realistic Clothing Modeling (ECCV 2018 · similarity 0.89)
- SCULPT: Shape-Conditioned Unpaired Learning of Pose-dependent Clothed and Textured Human Meshes (CVPR 2024 · similarity 0.89)
- DressWild: Feed-Forward Pose-Agnostic Garment Sewing Pattern Generation from In-the-Wild Images (NeurIPS 2026 · similarity 0.89)
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.
