首页|New Androids Data Have Been Reported by Researchers at Agency for Science (ExTra CT - Explainable trajectory corrections for language-based human-robot interacti on using textual feature descriptions)
New Androids Data Have Been Reported by Researchers at Agency for Science (ExTra CT - Explainable trajectory corrections for language-based human-robot interacti on using textual feature descriptions)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
Current study results on androids have been published. According to news reporting from Singapore, Singapore, by NewsR x journalists, research stated, "IntroductionIn human-robot interaction (HRI), u nderstanding human intent is crucial for robots to perform tasks that align with user preferences." Our news reporters obtained a quote from the research from Agency for Science: " Traditional methods that aim to modify robot trajectories based on language corr ections often require extensive training to generalize across diverse objects, i nitial trajectories, and scenarios. This work presents ExTraCT, a modular framew ork designed to modify robot trajectories (and behaviour) using natural language input. MethodsUnlike traditional end-to-end learning approaches, ExTraCT separa tes language understanding from trajectory modification, allowing robots to adap t language corrections to new tasks-including those with complex motions like sc ooping-as well as various initial trajectories and object configurations without additional end-to-end training. ExTraCT leverages Large Language Models (LLMs) to semantically match language corrections to predefined trajectory modification functions, allowing the robot to make necessary adjustments to its path. This m odular approach overcomes the limitations of pre-trained datasets and offers ver satility across various applications. ResultsComprehensive user studies conducte d in simulation and with a physical robot arm demonstrated that ExTraCT's trajec tory corrections are more accurate and preferred by users in 80% o f cases compared to the baseline."
Agency for ScienceSingaporeSingaporeAsiaEmerging TechnologiesHuman-Robot InteractionMachine LearningNano-r obotRobotRobotics
NETL
NSTL
