GAS-Robo: Converting Scene Into a Grid-Action Space for LLM-Driven Open-Ended Robotic Manipulation
- 1 October 2026
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 11 (10) , 12096-12103
Abstract
No abstract availableFunding Information
- Beijing Municipal Natural Science Foundation (L257009)
- National Natural Science Foundation of China (62495082)
This publication has 20 references indexed in Scilit:
- OmniManip: Towards General Robotic Manipulation via Object-Centric Interaction Primitives as Spatial Constraints2025
- Robotic Visual Instruction2025
- A Real-to-Sim-to-Real Approach to Robotic Manipulation with VLM-Generated Iterative Keypoint Rewards2025
- KALIE: Fine-Tuning Vision-Language Models for Open-World Manipulation Without Robot Data2025
- KALM: Keypoint Abstraction Using Large Models for Object-Relative Imitation Learning2025
- Grounding DINO: Marrying DINO with Grounded Pre-training for Open-Set Object Detection2024
- Segment Anything2023
- Code as Policies: Language Model Programs for Embodied Control2023
- Coarse-to-Fine Q-attention: Efficient Learning for Visual Robotic Manipulation via Discretisation2022
- Deep Imitation Learning for Complex Manipulation Tasks from Virtual Reality Teleoperation2018