Patents Examined by Gertrude Arther Jeanglaude
  • Patent number: 10203210
    Abstract: System, methods, and other embodiments described herein relate to identifying changes to a location. In one embodiment, a method includes, in response to receiving current sensor data about the location, extracting a semantic feature distribution from the current sensor data according to a semantic-change model that identifies sensed integral features and sensed transient features from the current sensor data. The method includes identifying a scene change to the location by determining whether the sensed integral features differ from learned integral features of a normal distribution that indicates both learned integral features and learned transient features of the location. The method includes updating a map of the location in response to identifying that one or more of the sensed integral features differ from the learned integral features of the normal distribution. The method includes controlling one or more vehicle systems of the vehicle according to the map.
    Type: Grant
    Filed: November 3, 2017
    Date of Patent: February 12, 2019
    Assignee: Toyota Research Institute, Inc.
    Inventor: Takaaki Tagawa