Research map / 3D vision / Gaussian splatting
CloDS: Visual-Only Unsupervised Cloth Dynamics Learning in Unknown Conditions
Yu-Liang Zhan, Jian Li, Wenbing Huang, Yang Liu, Hao Sun
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 |
|---|---|---|---|
| train CloDS on videos with 50 extra video trajectories on top of MGN* mesh pretraining | CloDS (MGN) · FLAGSIMPLE unviewed interpolation | rollout RMSE (Avg.): 0.1460 -> 0.1399 | Improved |
| train on all videos (CloDS**) versus MGN with full mesh supervision | CloDS · FLAGSIMPLE viewed interpolation | rollout RMSE (Avg.): 0.1286 (MGN) vs 0.1294 (CloDS**) | No clear effect |
| use fewer videos (CloDS* 50 videos vs CloDS 50 videos plus meshes) for unviewed trajectories | CloDS* · FLAGSIMPLE unviewed interpolation | rollout RMSE (Avg.): 0.1399 (CloDS) vs 0.1428 (CloDS*) | Hurt |
| replace GaMeS fixed opacity with dual-position opacity modulation (SMGS) (ablation) | mesh Gaussian splatting · Dynamic scene novel view synthesis | PSNR (dB): 33.0249 -> 36.2368 | Improved |
| add world-space and mesh-space coordinates to opacity in mesh extraction (ablation) | SMGS · Mesh extraction stage ablation | mesh error over time | Improved |
| render videos through 3D mesh prediction instead of 2D video prediction (CloDS vs SimVP) | CloDS · Cloth video prediction | PSNR (dB): 25.4770 -> 26.6207 | Improved |
| train on multiple trajectories and test on unseen ones (CloDS vs SimVP) | CloDS · Cloth video prediction (generalization) | PSNR (dB): 14.2391 -> 25.2380 | Improved |
| swap MGN for HCMT, DHMP or BSMS-GNN backbone and add video training (ablation) | CloDS · Neural simulator ablation | rollout RMSE | Improved |
| add noise to initial mesh (Gaussian, translation, scaling) (ablation) | CloDS* · FLAGSIMPLE unviewed interpolation | rollout RMSE d_avg t<=300: 0.1437 -> 0.1495 (scaling noise) | Hurt |
| add Gaussian noise to initial mesh (ablation) | CloDS* · FLAGSIMPLE viewed interpolation | rollout RMSE d_avg t<=300: 0.1313 -> 0.1362 | Hurt |
Closest papers by meaning
- Dynamic Texture Modeling of 3D Clothed Gaussian Avatars from a Single Video (ICLR 2026 · similarity 0.88)
- Learning 3D Garment Animation from Trajectories of A Piece of Cloth (NeurIPS 2024 · similarity 0.87)
- SNUG: Self-Supervised Neural Dynamic Garments (CVPR 2022 · similarity 0.87)
- GarmentPile: Point-Level Visual Affordance Guided Retrieval and Adaptation for Cluttered Garments Manipulation (CVPR 2025 · similarity 0.86)
- NeuralClothSim: Neural Deformation Fields Meet the Thin Shell Theory (NeurIPS 2024 · similarity 0.86)
- Motion Decoupled 3D Gaussian Splatting for Dynamic Object Representation (AAAI 2025 · similarity 0.85)
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.
