×

Semantic parsing of objects in video

  • US 9,679,201 B2
  • Filed: 01/18/2016
  • Issued: 06/13/2017
  • Est. Priority Date: 07/28/2010
  • Status: Active Grant
First Claim
Patent Images

1. A method comprising:

  • producing a plurality of versions of an image of an object derived from an input, wherein each version comprises one of one multiple resolutions of said image of said object;

    computing an appearance probability at each of a plurality of regions on the one or more lowest resolution versions of said plurality of versions of said image for at least one attribute for said object;

    computing a resolution context score for each of the plurality of regions in the one or more lowest resolution versions, wherein the resolution context score comprises a weighted average computed from a plurality of scores for a next higher resolution version of said image;

    computing a geometric score for each region of said plurality of regions in the one or more lowest resolution versions, said geometric score computing a probability of a region matching stored reference data for a reference object corresponding to the detected object with respect to angles and distances among the plurality of regions;

    determining a configuration of the at least one attribute in the one or more lowest resolution versions based on at least the appearance probability, the resolution context score, and the geometric score in each of the plurality of regions; and

    outputting said configuration.

View all claims
  • 2 Assignments
Timeline View
Assignment View
    ×
    ×