首页|Researchers at University of Illinois Release New Data on Robotics (W-rizz: a We akly-supervised Framework for Relative Traversability Estimation In Mobile Robot ics)
Researchers at University of Illinois Release New Data on Robotics (W-rizz: a We akly-supervised Framework for Relative Traversability Estimation In Mobile Robot ics)
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 – Investigators publish new report on Ro botics. According to news reporting originating in Champaign, Illinois, by NewsR x journalists, research stated, “Successful deployment of mobile robots in unstr uctured domains requires an understanding of the environment and terrain to avoi d hazardous areas, getting stuck, and colliding with obstacles. Traversability e stimation-which predicts where in the environment a robot can travel-is one prom inent approach that tackles this problem.” Financial support for this research came from National Robotics Initiative 2.0. The news reporters obtained a quote from the research from the University of Ill inois, “Existing geometric methods may ignore important semantic considerations, while semantic segmentation approaches involve a tedious labeling process. Rece nt self-supervised methods reduce labeling tedium, but require additional data o r models and tend to struggle to explicitly label untraversable areas. To addres s these limitations, we introduce a weakly-supervised method for relative traver sability estimation. Our method involves manually annotating the relative traver sability of a small number of point pairs, which significantly reduces labeling effort compared to traditional segmentation-based methods and avoids the limitat ions of self-supervised methods. We further improve the performance of our metho d through a novel cross-image labeling strategy and loss function.”
ChampaignIllinoisUnited StatesNort h and Central AmericaEmerging TechnologiesMachine LearningRobotRoboticsUniversity of Illinois
NETL
NSTL
