用少量尝试教会机器人抓握，适配多种手型
AnyDexGrasp: General Dexterous Grasping for Different Hands with Human-level Learning Efficiency
- 分两阶段学习：先生成通用接触表征，再为每种手训练专属抓取策略
- 仅需数百次尝试，在40个物体上实现75%-95%抓取成功率
- 适合需要灵活抓握的人形机器人、假肢等实际场景
我们提出一种高效学习方法，仅需少量数据即可实现不同机械手的灵巧抓握，显著提升机器人操作能力。与传统方法需数百万抓取标签不同，本方法仅需在40个训练物体上进行数百次真实抓取尝试，便达到人类级学习效率。该方法将抓取过程分为两个阶段：首先，一个通用模型将场景几何映射为不依赖具体手型的中间接触中心表征；其次，针对每种机械手单独训练一个抓取决策模型，将这些表征转化为最终抓取姿态。实测结果表明，在包含超过150种新物体的复杂环境中，三种不同机械手的抓取成功率达75%-95%，随着训练物体增加，成功率提升至80%-98%。该方法展现出在人形机器人、假肢及其他需要强鲁棒性、高灵活性操作的领域的广阔应用前景。
原文摘要 · Abstract (English)
We introduce an efficient approach for learning dexterous grasping with minimal data, advancing robotic manipulation capabilities across different robotic hands. Unlike traditional methods that require millions of grasp labels for each robotic hand, our method achieves high performance with human-level learning efficiency: only hundreds of grasp attempts on 40 training objects. The approach separates the grasping process into two stages: first, a universal model maps scene geometry to intermediate contact-centric grasp representations, independent of specific robotic hands. Next, a unique grasp decision model is trained for each robotic hand through real-world trial and error, translating these representations into final grasp poses. Our results show a grasp success rate of 75-95\% across three different robotic hands in real-world cluttered environments with over 150 novel objects, improving to 80-98\% with increased training objects. This adaptable method demonstrates promising applications for humanoid robots, prosthetics, and other domains requiring robust, versatile robotic manipulation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。