Research map / Vision-language and multimodal / Multimodal LLMs
GraspCoT: Integrating Physical Property Reasoning for 6-DoF Grasping under Flexible Language Instructions
Xiaomeng Chu, Jiajun Deng, Guoliang You, Wei Liu, Xingchen Li, Jianmin Ji, Yanyong Zhang
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 physical-property CoT reasoning (ablation) | GraspCoT · IntentGrasp | CR@0.2: 0.4164 -> 0.5587 | Improved |
| add physical-property CoT reasoning (ablation) | GraspCoT · IntentGrasp | EMD: 0.2878 -> 0.2520 | Improved |
| add material property analysis (ablation) | GraspCoT · IntentGrasp | CR@0.2: 0.4164 -> 0.5091 | Improved |
| add surface property analysis (ablation) | GraspCoT · IntentGrasp | CR@0.2: 0.4164 -> 0.4818 | Improved |
| add shape property analysis (ablation) | GraspCoT · IntentGrasp | CR@0.2: 0.4164 -> 0.4716 | Improved |
| add shape analysis to material and surface (ablation) | GraspCoT · IntentGrasp | CR@0.2: 0.5177 -> 0.5386 | Improved |
| add grasp action selection (ablation) | GraspCoT · IntentGrasp | CR@0.2: 0.5386 -> 0.5587 | Improved |
| add grasp action selection (ablation) | GraspCoT · IntentGrasp | EMD: 0.2572 -> 0.2520 | Improved |
| increase virtual views from 3 to 4 (ablation) | GraspCoT · IntentGrasp | CR@0.2: 0.4962 -> 0.5587 | Improved |
| increase virtual views from 4 to 5 (ablation) | GraspCoT · IntentGrasp | CR@0.2: 0.5587 -> 0.5581 | No clear effect |
| add CoT in multi-target scenarios (ablation) | GraspCoT · IntentGrasp (multiple targets) | CR@0.2: 0.4183 -> 0.5571 | Improved |
| add CoT in single-target scenarios (ablation) | GraspCoT · IntentGrasp (single target) | CR@0.2: 0.4108 -> 0.5590 | Improved |
Closest papers by meaning
- Language-driven Grasp Detection (CVPR 2024 · similarity 0.89)
- Generalizing 6-DoF Grasp Detection via Domain Prior Knowledge (CVPR 2024 · similarity 0.88)
- RealVLG-R1: A Large-Scale Real-World Visual-Language Grounding Benchmark for Robotic Perception and Manipulation (CVPR 2026 · similarity 0.88)
- OmniManip: Towards General Robotic Manipulation via Object-Centric Interaction Primitives as Spatial Constraints (CVPR 2025 · similarity 0.88)
- PAC Bench: Do Foundation Models Understand Prerequisites for Executing Manipulation Policies? (NeurIPS 2025 · similarity 0.87)
- ManipLLM: Embodied Multimodal Large Language Model for Object-Centric Robotic Manipulation (CVPR 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.
