通用机器人控制中的模仿学习技术  

Imitation Learning Technology in Universal Robot Control

作　　者：庞浩 PANG Hao(Independent Variable Robotics Technology(Shenzhen)Co.,Ltd.,Shenzhen,Guangdong 518000,China)

机构地区：[1]自变量机器人科技(深圳)有限公司,广东深圳518100

出　　处：《移动信息》2024年第10期368-370,共3页Mobile Information

摘　　要：在具身智能(Embodied AI)领域,模仿学习是赋予机器人各种复杂操作能力的一种有效方法。其能通过各种各样的技术手段获取人类的离线和在线操作数据,为构建通用机器人控制模型提供重要支持。基于这些操作数据的模仿学习技术,已成为具身智能领域的重要研究方向。In the field of embodied intelligence(Embodied AI),imitation learning is an effective method to endow robots with various complex operation abilities.It can obtain human offline and online operation data through various technical means,which can provide important support for building general robot control models.Imitation learning technology based on these operation data has become an important research direction in the field of embodied intelligence.

关 键 词：通用机器人 具身智能 通用人工智能 模仿学习 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]