Research map / 3D vision / 3D generation
RAGrasp: A Retrieval-Augmented Framework with Diversity-Aware Modeling for Dexterous Grasp Generation
Zhuo Yang, Sanping Zhou, Sen Wang, Jingyi Tian, lijiayi lijiayi, Gang Hua, Le Wang
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 geometry-aware hand pose initialization via retrieval (GAI) (ablation) | mobius normalizing flow grasp generator · Dexonomy (type-unconditioned) | avg success: 34.2 -> 39.1 | Improved |
| add point cloud completion (ablation) | normalizing flow baseline · Dexonomy (type-unconditioned) | avg success: 34.2 -> 36.0 | Improved |
| retrieve dataset ground-truth grasps instead of model-generated grasps (ablation) | Base+Comp · Dexonomy (type-unconditioned) | avg success: 36.0 -> 30.4 | Hurt |
| build retrieval dataset from model-generated successful grasps (data collection module) (ablation) | Base+Comp+Ret · Dexonomy (type-unconditioned) | avg success: 30.4 -> 39.1 | Improved |
| RAGrasp with GAI on fingertip grasps | RAGrasp vs UDG · BODex | GSR: 80.1 -> 86.7 | Improved |
| RAGrasp vs UDG on diversity metric | RAGrasp · BODex | Diversity: 33.4 -> 37.7 | Mixed |
| use high DAS threshold (success-favoring) | RAGrasp · Dexonomy (type-conditioned) | GSR: 63.9 -> 71.5 | Improved |
| use low DAS threshold (diversity-favoring) | RAGrasp · Dexonomy (type-conditioned) | GSR: 63.9 -> 65.6 | Improved |
| lower DAS threshold to activate more grasp types (ablation) | DAS · Dexonomy | diversity vs success | Mixed |
Closest papers by meaning
- RAGG: Retrieval-Augmented Grasp Generation Model (AAAI 2025 · similarity 0.91)
- Generalizing 6-DoF Grasp Detection via Domain Prior Knowledge (CVPR 2024 · similarity 0.88)
- Contact Map Transfer with Conditional Diffusion Model for Generalizable Dexterous Grasp Generation (NeurIPS 2025 · similarity 0.88)
- DexGrasp Anything: Towards Universal Robotic Dexterous Grasping with Physics Awareness (CVPR 2025 · similarity 0.88)
- Dexterous Grasp Transformer (CVPR 2024 · similarity 0.87)
- CADGrasp: Learning Contact and Collision Aware General Dexterous Grasping in Cluttered Scenes (NeurIPS 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.
