使用图像和文本对多种任务和机器人进行基于 CLIP 特征的随机控制

IF 1.4 4区计算机科学 Q4 ROBOTICS

Advanced Robotics Pub Date : 2024-08-01 DOI:10.1080/01691864.2024.2379381

Kazuki Shibata, Hideki Deguchi, Shun Taguchi

引用次数: 0

摘要

本研究提出了一种利用视觉语言模型（VLM）的控制框架，适用于多种任务和机器人。值得注意的是，现有的使用视觉语言模型的控制方法已在各种任务和机器人中实现了高性能。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

CLIP feature-based randomized control using images and text for multiple tasks and robots

This study presents a control framework leveraging vision language models (VLMs) for multiple tasks and robots. Notably, existing control methods using VLMs have achieved high performance in variou...

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Advanced Robotics 工程技术-机器人学

CiteScore

4.10

自引率

20.00%

发文量

102

审稿时长

5.3 months

期刊介绍： Advanced Robotics (AR) is the international journal of the Robotics Society of Japan and has a history of more than twenty years. It is an interdisciplinary journal which integrates publication of all aspects of research on robotics science and technology. Advanced Robotics publishes original research papers and survey papers from all over the world. Issues contain papers on analysis, theory, design, development, implementation and use of robots and robot technology. The journal covers both fundamental robotics and robotics related to applied fields such as service robotics, field robotics, medical robotics, rescue robotics, space robotics, underwater robotics, agriculture robotics, industrial robotics, and robots in emerging fields. It also covers aspects of social and managerial analysis and policy regarding robots. Advanced Robotics (AR) is an international, ranked, peer-reviewed journal which publishes original research contributions to scientific knowledge. All manuscript submissions are subject to initial appraisal by the Editor, and, if found suitable for further consideration, to peer review by independent, anonymous expert referees.