arXivDaily arXiv每日学术速递 周一至周五更新
​通知:联邦节假日调休:哥伦布日 / 原住民日(Columbus Day / Indigenous Peoples' Day),今日(13日)无更新。
arXiv 2502.04873cs.RO

无需训练的任务导向抓取生成

Training-free Task-oriented Grasp Generation

Jiaming Wang, Diwen Liu, Jizhuo Chen, Harold Soh

AI总结:

本研究提出一种无需训练的任务导向抓取生成流水线,结合预训练抓取模型与视觉语言模型,通过五种利用不同抓取视觉表示的查询策略,在抓取成功率和任务符合率上显著优于基线,为机器人操作相关研究提供新思路。

摘要:

本文提出了一种无需训练的任务导向抓取生成流水线,该流水线将预训练抓取生成模型与视觉语言模型(vision-language models, VLMs)相结合。与仅关注稳定抓取的传统方法不同,我们的方法利用VLMs的语义推理能力纳入了任务特定要求。我们评估了五种查询策略,每种策略都使用候选抓取的不同视觉表示,并证明其在抓取成功率和任务符合率上均显著优于基线方法,总体成功率的绝对提升最高达36.9%。我们的结果凸显了VLMs增强任务导向操作的潜力,为机器人抓取与人机交互的未来研究提供了见解。

英文摘要:

This paper presents a training-free pipeline for task-oriented grasp generation that combines pre-trained grasp generation models with vision-language models (VLMs). Unlike traditional approaches that focus solely on stable grasps, our method incorporates task-specific requirements by leveraging the semantic reasoning capabilities of VLMs. We evaluate five querying strategies, each utilizing different visual representations of candidate grasps, and demonstrate significant improvements over a baseline method in both grasp success and task compliance rates, with absolute gains of up to 36.9\% in overall success rate. Our results underline the potential of VLMs to enhance task-oriented manipulation, providing insights for future research in robotic grasping and human-robot interaction.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑