Investigators from National University of Defense Technology Have Reported New Data on Robotics and Automation (Ilabel: Revealing Objects In Neural Fields)

摘要

By a News Reporter-Staff News Editor at Robotics Machine Learning DailyNews Daily News - Investigators publish new report on Robotics - Robotics and Automation. According tonews originating from Changsha, People’s Republic of China, by NewsRx correspondents, research stated,“A neural field trained with self-supervision to efficiently represent the geometry and colour of a 3D scenetends to automatically decompose it into coherent and accurate object-like regions, which can be revealedwith sparse labelling interactions to produce a 3D semantic scene segmentation. Our real-time iLabelsystem takes input from a hand-held RGB-D camera, requires zero prior training data, and works in an‘open set’ manner, with semantic classes defined on the fly by the user. iLabel’s underlying model is asimple multilayer perceptron (MLP), trained from scratch to learn a neural representation of a single 3Dscene.”