Výsledky vyhledávání

Report

Data Scaling Laws in Imitation Learning for Robotic Manipulation

Autor: Lin, Fanqi, Hu, Yingdong, Sheng, Pingyue, Wen, Chuan, You, Jiacheng, Gao, Yang

Data scaling has revolutionized fields like natural language processing and computer vision, providing models with remarkable generalization capabilities. In this paper, we investigate whether similar data scaling laws exist in robotics, particularly

Externí odkaz: http://arxiv.org/abs/2410.18647

Zobrazit plný text záznamu

Report

CoPa: General Robotic Manipulation through Spatial Constraints of Parts with Foundation Models

Autor: Huang, Haoxu, Lin, Fanqi, Hu, Yingdong, Wang, Shengjie, Gao, Yang

Foundation models pre-trained on web-scale data are shown to encapsulate extensive world knowledge beneficial for robotic manipulation in the form of task planning. However, the actual physical implementation of these plans often relies on task-speci

Externí odkaz: http://arxiv.org/abs/2403.08248

Zobrazit plný text záznamu

Report

Look Before You Leap: Unveiling the Power of GPT-4V in Robotic Vision-Language Planning

Autor: Hu, Yingdong, Lin, Fanqi, Zhang, Tong, Yi, Li, Gao, Yang

In this study, we are interested in imbuing robots with the capability of physically-grounded task planning. Recent advancements have shown that large language models (LLMs) possess extensive knowledge useful in robotic tasks, especially in reasoning

Externí odkaz: http://arxiv.org/abs/2311.17842

Zobrazit plný text záznamu

Report

Diverse Policies Converge in Reward-free Markov Decision Processe

Autor: Lin, Fanqi, Huang, Shiyu, Tu, Weiwei

Reinforcement learning has achieved great success in many decision-making tasks, and traditional reinforcement learning algorithms are mainly designed for obtaining a single optimal solution. However, recent works show the importance of developing di

Externí odkaz: http://arxiv.org/abs/2308.11924

Zobrazit plný text záznamu

Report

TiZero: Mastering Multi-Agent Football with Curriculum Learning and Self-Play

Autor: Lin, Fanqi, Huang, Shiyu, Pearce, Tim, Chen, Wenze, Tu, Wei-Wei

Multi-agent football poses an unsolved challenge in AI research. Existing work has focused on tackling simplified scenarios of the game, or else leveraging expert demonstrations. In this paper, we develop a multi-agent system to play the full 11 vs.

Externí odkaz: http://arxiv.org/abs/2302.07515

Zobrazit plný text záznamu