Research map / 3D vision / Human pose estimation
Learning Part-Aware Dense 3D Feature Field For Generalizable Articulated Object Manipulation
Yue Chen, Muqing Jiang, Kaifeng Zheng, Jiaqi Liang, Chenrui Tie, Haoran Lu, Ruihai Wu, Hao Dong
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 stacking additional transformer blocks and remove PTv3 downsampling (ablation) | PADP · Put in Drawer / Turn Tap / Open Box | success rate (%): 58 -> 62; 63 -> 69; 59 -> 66 | Improved |
| add geometric contrastive loss (ablation) | PA3FF · Put in Drawer / Turn Tap / Open Box | success rate (%): 54 -> 62; 57 -> 69; 55 -> 66 | Improved |
| add semantic loss (SigLIP part-name alignment) (ablation) | PA3FF · Put in Drawer / Turn Tap / Open Box | success rate (%): 46 -> 62; 53 -> 69; 52 -> 66 | Improved |
| replace frozen Sonata features with PA3FF in DP3-style policy (ablation) | DP3 · Put in Drawer | success rate (%): 39 -> 62 (Sonata+DP3 vs PADP) | Improved |
| add Sonata features to DP3 without modification (ablation) | DP3 · Put in Drawer | success rate (%): 37 -> 39 | Improved |
| use PADP instead of GenDP (ablation) | policy · PartInstruct average | success rate (%): 19.36 -> 28.79 | Improved |
| use PADP instead of GenDP under environment disturbance (ablation) | policy · Real-world Open Bottle | completion rate (%): 30 -> 60 | Improved |
Closest papers by meaning
- General Articulated Objects Manipulation in Real Images via Part-Aware Diffusion Process (NeurIPS 2024 · similarity 0.90)
- DenseMatcher: Learning 3D Semantic Correspondence for Category-Level Manipulation from a Single Demo (ICLR 2025 · similarity 0.89)
- SparseDFF: Sparse-View Feature Distillation for One-Shot Dexterous Manipulation (ICLR 2024 · similarity 0.88)
- G3Flow: Generative 3D Semantic Flow for Pose-aware and Generalizable Object Manipulation (CVPR 2025 · similarity 0.88)
- GravMAD: Grounded Spatial Value Maps Guided Action Diffusion for Generalized 3D Manipulation (ICLR 2025 · similarity 0.87)
- Lift3D Policy: Lifting 2D Foundation Models for Robust 3D Robotic Manipulation (CVPR 2025 · 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.
