US 11,810,597 B2
Video ingestion and clip creation
Matthew Alan Townsend, Redwood City, CA (US); Rohith Mysore Vijaya Kumar, Sunnyvale, CA (US); Yadunandana Nagaraja Rao, Sunnyvale, CA (US); Ambrish Tyagi, Palo Alto, CA (US); Eduard Oks, Holon (IL); and Apoorv Chaudhri, Sunnyvale, CA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Oct. 4, 2021, as Appl. No. 17/492,781.
Application 17/492,781 is a continuation of application No. 14/870,227, filed on Sep. 30, 2015, granted, now 11,158,344.
Prior Publication US 2022/0122639 A1, Apr. 21, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G11B 27/031 (2006.01); H04N 9/79 (2006.01)
CPC G11B 27/031 (2013.01) [H04N 9/79 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method, comprising:
receiving first video data;
receiving second video data;
determining a first theme score associated with a first candidate theme, the first theme score representing a likelihood that the first candidate theme corresponds to both the first video data and the second video data;
determining a second theme score associated with a second candidate theme, the second theme score representing a likelihood that the second candidate theme corresponds to both the first video data and the second video data;
based at least in part on the first theme score and the second theme score, selecting the first candidate theme, instead of the second candidate theme, as a theme;
based on selecting the first candidate theme as the theme, determining a first group of video clips associated with at least a portion of the first video data and at least a portion of the second video data; and
generating a video summarization using the theme and the first group of video clips.