首页|New Data from Jozef Stefan Institute Illuminate Findings in Robotics (Simulation -aided Handover Prediction From Video Using Recurrent Image-to-motion Networks)
New Data from Jozef Stefan Institute Illuminate Findings in Robotics (Simulation -aided Handover Prediction From Video Using Recurrent Image-to-motion Networks)
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
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Data detailed on Robotics have been pr esented. According to news reporting originating from Ljubljana, Slovenia, by Ne wsRx correspondents, research stated, "Recent advances in deep neural networks h ave opened up new possibilities for visuomotor robot learning. In the context of human-robot or robot-robot collaboration, such networks can be trained to predi ct future poses and this information can be used to improve the dynamics of coop erative tasks." Financial supporters for this research include Slovenian Research Agency - Slove nia, European Union (EU), New Energy and Industrial Technology Development Organ ization (NEDO), Grants-in-Aid for Scientific Research (KAKENHI), Japan Science a nd Technology Agency (JST) Mirai Program, Tateishi Science and Technology Founda tion. Our news editors obtained a quote from the research from Jozef Stefan Institute, "This is important, both in terms of realizing various cooperative behaviors, a nd for ensuring safety. In this article, we propose a recurrent neural architect ure, capable of transforming variable-length input motion videos into a set of p arameters describing a robot trajectory, where predictions can be made after rec eiving only a few frames. A simulation environment is utilized to expand the tra ining database and to improve generalization capability of the network. The resu lting architecture demonstrates good accuracy when predicting handover trajector ies, with models trained on synthetic and real data showing better performance t han when trained on real or simulated data only."
LjubljanaSloveniaEuropeEmerging Te chnologiesMachine LearningNano-robotRobotRoboticsJozef Stefan Institut e
NETL
NSTL
