US 12,132,972 B2
Methods and systems for detecting content within media streams
Kiran Sanjeevan Cabeza, San Francisco, CA (US)
Assigned to VIZIO, INC., Irvine, CA (US)
Filed by VIZIO, INC., Irvine, CA (US)
Filed on Sep. 6, 2022, as Appl. No. 17/903,754.
Claims priority of provisional application 63/278,415, filed on Nov. 11, 2021.
Claims priority of provisional application 63/241,446, filed on Sep. 7, 2021.
Prior Publication US 2023/0070957 A1, Mar. 9, 2023
Int. Cl. H04N 21/8352 (2011.01); H04N 21/845 (2011.01)
CPC H04N 21/8456 (2013.01) [H04N 21/8352 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A method comprising:
receiving a media stream;
defining, from the media stream, a set of media segments, each media segment representing a portion of the media stream;
executing a first machine-learning model configured to identify a first media segment of the set of media segments representative of a first boundary and a second media segment of the set of media segments representative of a second boundary, wherein the first boundary is indicative of a potential content transition;
executing, for at least one media segment of the set of media segments that is between the first media segment and the second media segment, a second machine-learning model configured to generate a prediction corresponding to whether the at least one media segment corresponds to a particular content type;
determining that a subset of the set of media segments that includes media segments between the first media segment and the second media segment correspond to the particular content type based on the first boundary, the second boundary, and the prediction; and
facilitating a transmission including an indication that the subset of the set of media segments corresponds to the particular content type.