Research map / Vision-language and multimodal / Robot action and navigation
From Seeing to Doing: Bridging Reasoning and Decision for Robotic Manipulation
Yifu Yuan, Haiqin Cui, Yibin Chen, Zibin Dong, Fei Ni, Longxin Kou, Jinyi Liu, Pengyi Li, YAN ZHENG, Jianye HAO
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 SrCoT spatial-relationship chain-of-thought reasoning (ablation) | FSD · VABench-Point | accuracy: 26.21 -> 61.82 | Improved |
| add self-consistency alignment (inverse understanding task) (ablation) | FSD · VABench-Point | accuracy: 55.92 -> 61.82 | Improved |
| add SrCoT reasoning (ablation) | FSD · VABench-VisualTrace | RMSE: 99.53 -> 78.26 | Improved |
| add SrCoT reasoning (ablation) | FSD · VABench-VisualTrace | LLM Score: 5.07 -> 6.21 | Improved |
| add self-consistency alignment (ablation) | FSD · VABench-VisualTrace | RMSE: 80.48 -> 78.26 | Improved |
| add self-consistency alignment (ablation) | FSD · VABench-VisualTrace | LLM Score: 5.92 -> 6.21 | Improved |
| train on weak-to-strong spatial data on top of ASMv2 | ASMv2-13B · EmbSpatial-Bench Test | accuracy: 57.4 -> 63.3 | Improved |
| use SrCoT-based FSD instead of RoboPoint for free-space reference | FSD vs RoboPoint · Where2Place | accuracy: 46.0 vs 45.8 | No clear effect |
| reasoning-based FSD instead of end-to-end DINOv2 Predictor trained on same trace data | visual trace prediction · VABench-VisualTrace | RMSE: 128.32 -> 78.26 | Improved |
| zero-shot FSD vs RoboPoint zero-shot | WidowX robot · SimplerEnv WidowX | average success rate (%): 17.7 -> 40.6 | Improved |
| zero-shot FSD vs fine-tuned pi0-fast | WidowX robot · SimplerEnv WidowX | average success rate (%): 48.3 vs 40.6 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- RoboBrain — RoboBrain: A Unified Brain Model for Robotic Manipulation from Abstract to Concrete (CVPR 2025) · better on 1 of 1 shared result
- MOKA — MokA: Multimodal Low-Rank Adaptation for MLLMs (NeurIPS 2025) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- OmniManip: Towards General Robotic Manipulation via Object-Centric Interaction Primitives as Spatial Constraints (CVPR 2025 · similarity 0.92)
- Seeing Across Views: Benchmarking Spatial Reasoning of Vision-Language Models in Robotic Scenes (ICLR 2026 · similarity 0.92)
- RoboProcessBench: Benchmarking Process-Aware Understanding in Visual-Language Robotic Manipulation (NeurIPS 2026 · similarity 0.91)
- Localizing, Structuring, and Rendering: Bridging 3D and 2D Vision-Language-Action Models for Robotic Manipulation (CVPR 2026 · similarity 0.91)
- Exploring the Limits of Vision-Language-Action Manipulation in Cross-task Generalization (NeurIPS 2025 · similarity 0.91)
- Robotic Visual Instruction (CVPR 2025 · similarity 0.91)
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.
