Research map / 3D vision / Face and shape
Analyzing Clothing Layer Deformation Statistics of 3D Human Motions
Jinlong Yang, Jean-Sebastien Franco, Franck Hetroy-Wheeler, Stefanie Wuhrer
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 |
|---|---|---|---|
| keep 40 principal components instead of 5 for offset clothing layer reconstruction (ablation) | PCA clothing layer model · PCA reconstruction, Bouncing sequence | reconstruction error (mm): 22.52 -> 6.49 | Improved |
| keep 40 principal components instead of 5 (ablation) | PCA clothing layer model · PCA reconstruction, Squat 1 sequence | reconstruction error (mm): 18.94 -> 2.67 | Improved |
| keep 40 principal components instead of 1 (ablation) | PCA clothing layer model · PCA reconstruction, Squat 1 sequence | reconstruction error (mm): 46.33 -> 2.67 | Improved |
| keep all principal components instead of 40 (ablation) | PCA clothing layer model · PCA reconstruction, Bouncing sequence | reconstruction error (mm): 6.49 -> 0.10 | Improved |
| use more principal components (ablation) | PCA clothing layer model · PCA reconstruction, 10 sequences | average reconstruction error | Improved |
| use fuzzy association with higher nf instead of nf = 1 (ablation) | offset clothing layer extraction · skirt sequence | seam artifacts (qualitative) | Improved |
| regress PCA coefficients from motion with a neural network, tested on held-out frames | clothing regression · Crane sequence | prediction error (mm): 4.71 -> 4.28 | No clear effect |
| evaluate regression on held-out frames versus training frames | clothing regression · Samba sequence | error (mm): 4.27 -> 7.44 | Hurt |
| add Gaussian noise to material and size parameters during regression training | material and fit regression · material and clothing fit synthesis | overfitting to few training points (qualitative) | Improved |
| reduce material parameters from 39 to 4 dimensions and deformation space with PCA | material regression · simulated garment, held-out material | wrinkle detail (qualitative) | Hurt |
Closest papers by meaning
- TailorNet: Predicting Clothing in 3D as a Function of Human Pose, Shape and Garment Style (CVPR 2020 · similarity 0.90)
- SCALE: Modeling Clothed Humans with a Surface Codec of Articulated Local Elements (CVPR 2021 · similarity 0.90)
- SNUG: Self-Supervised Neural Dynamic Garments (CVPR 2022 · similarity 0.90)
- Garment4D: Garment Reconstruction from Point Cloud Sequences (NeurIPS 2021 · similarity 0.89)
- 4D-DRESS: A 4D Dataset of Real-World Human Clothing With Semantic Annotations (CVPR 2024 · similarity 0.88)
- Learning 3D Garment Animation from Trajectories of A Piece of Cloth (NeurIPS 2024 · 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.
