US 12,301,897 B2
Emotion evaluation of contents
Ronica Jethwa, Mountain View, CA (US); Nam Vo, San Jose, CA (US); Fei Xiao, San Jose, CA (US); and Abhishek Bambha, Burlingame, CA (US)
Assigned to Roku, Inc., San Jose, CA (US)
Filed by Roku, Inc., San Jose, CA (US)
Filed on Jan. 29, 2024, as Appl. No. 18/425,803.
Application 18/425,803 is a continuation of application No. 17/877,124, filed on Jul. 29, 2022, granted, now 11,930,226.
Prior Publication US 2024/0171783 A1, May 23, 2024
Int. Cl. H04N 21/234 (2011.01); G06V 20/40 (2022.01); H04N 21/25 (2011.01); H04N 21/81 (2011.01); H04N 21/8549 (2011.01)
CPC H04N 21/23418 (2013.01) [G06V 20/41 (2022.01); H04N 21/251 (2013.01); H04N 21/812 (2013.01); H04N 21/8549 (2013.01); G06V 2201/10 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method, comprising:
processing, by at least one computer processor, content to obtain at least one of scene metadata associated with one or more scenes of the content, frame metadata associated with one or more frames of the one or more scenes, or content metadata associated with the content;
providing a representation of the at least one of the scene metadata, the frame metadata or the content metadata to a trained emotion model;
obtaining a frame emotion value for each of the one or more frames of the one or more scenes from the trained emotion model, the frame emotion value being based on the at least one of the scene metadata, the frame metadata or the content metadata;
generating a sequence of frame emotion values for each of the one or more scenes based on the one or more frames of the one or more scenes; and
generating at least one scene emotion value for the one or more scenes based on the sequence of frame emotion values.