首页|Investigators at Stanford University Describe Findings in Robotics and Automatio n (Prime: Scaffolding Manipulation Tasks With Behavior Primitives for Data-effic ient Imitation Learning)
Investigators at Stanford University Describe Findings in Robotics and Automatio n (Prime: Scaffolding Manipulation Tasks With Behavior Primitives for Data-effic ient Imitation Learning)
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
2024 OCT 03 (NewsRx)-By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Data detailed on Robotics - Robotics a nd Automation have been presented. According to news reporting from Stanford, Ca lifornia, by NewsRx journalists, research stated, "Imitation learning has shown great potential for enabling robots to acquire complex manipulation behaviors. H owever, these algorithms suffer from high sample complexity in long-horizon task s, where compounding errors accumulate over the task horizons." Financial supporters for this research include IIIS, Tsinghua University, Adviso r, Todor Stoyanov, Wallenberg AI, Autonomous Systems, and Software Program (WASP ), National Science Foundation (NSF), Office of Naval Research, UT Good Systems, Machine Learning Laboratory. The news correspondents obtained a quote from the research from Stanford Univers ity, "We present PRIME (PRimitive-based IMitation with data Efficiency), a behav ior primitive-based framework designed for improving the data efficiency of imit ation learning. PRIME scaffolds robot tasks by decomposing task demonstrations i nto primitive sequences, followed by learning a high-level control policy to seq uence primitives through imitation learning."
StanfordCaliforniaUnited StatesNor th and Central AmericaRobotics and AutomationRoboticsStanford University
NETL
NSTL
