Research map / Image generation / Autoregressive and latent generation
SNS-Grasp: Semantic-guided Noise Scaling for Grasp Generation
Zhenhua Tang, Yudian Zheng, Yuzhang Zhong, Haolun Li, Yanbin Hao, Chi-Man Pun
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 semantic-guided noise scaling (SNS3) to diffusion (ablation) | SNS-Diff · OakInk seen | Controllability Auc (%): 93.38 -> 97.74 | Improved |
| add SNS3 noise scaling (ablation) | SNS-Diff · OakInk seen | Penetration depth (cm): 0.608 -> 0.568 | Improved |
| scale noise by gradient norm only (SNS1) (ablation) | SNS-Diff · OakInk seen | Simulation displacement (cm): 1.141 -> 1.246 | Hurt |
| increase noise for critical joints (SNS2) instead of reducing it (ablation) | SNS-Diff · OakInk seen | Controllability Auc (%): 97.74 -> 96.48 | Hurt |
| use FGR refinement instead of GrabNet refine (ablation) | SNS-Grasp · OakInk seen | Simulation displacement (cm): 2.273 -> 1.411 | Improved |
| use FGR refinement instead of ProKAN (ablation) | SNS-Grasp · OakInk seen | Controllability Auc (%): 95.60 -> 97.39 | Improved |
| use FGR instead of ProKAN refinement (ablation) | SNS-Grasp · OakInk seen | Penetration depth (cm): 0.444 -> 0.446 | No clear effect |
| SNS-Grasp vs RAGG on unseen OakInk categories | SNS-Grasp · OakInk unseen | Intersection volume (cm3): RAGG 7.826 -> 6.933; GrabNet 6.228 lower than ours | Mixed |
| SNS-Grasp vs RAGG on GRAB | SNS-Grasp · GRAB | Controllability Auc (%): 62.11 -> 73.44 | Improved |
Closest papers by meaning
- RAGG: Retrieval-Augmented Grasp Generation Model (AAAI 2025 · similarity 0.88)
- Language-driven Grasp Detection (CVPR 2024 · similarity 0.87)
- Grasp2Grasp: Vision-Based Dexterous Grasp Translation via Schrödinger Bridges (NeurIPS 2025 · similarity 0.87)
- Grasp as You Say: Language-guided Dexterous Grasp Generation (NeurIPS 2024 · similarity 0.86)
- CoDA: Coordinated Diffusion Noise Optimization for Whole-Body Manipulation of Articulated Objects (NeurIPS 2025 · similarity 0.86)
- Contact Map Transfer with Conditional Diffusion Model for Generalizable Dexterous Grasp Generation (NeurIPS 2025 · similarity 0.86)
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.
