US 12,445,571 B1
Systems and methods for automatic aspect ratio adjustment for video content
Rohun Tripathi, Seattle, WA (US); and Baris Gecer, Post (GB)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on May 28, 2024, as Appl. No. 18/675,852.
Int. Cl. H04N 7/01 (2006.01); G06V 10/75 (2022.01); G06V 20/40 (2022.01); H04N 5/265 (2006.01)
CPC H04N 7/0122 (2013.01) [G06V 10/751 (2022.01); G06V 20/49 (2022.01); H04N 5/265 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
receiving first video content including a first plurality of video frames, wherein at least some of the first plurality of video frames are of a first aspect ratio;
separating, by one or more processors, the first video content into one or more segments, each of the one or more segments including a different camera cut of the first video content;
identifying, by the one or more processors, a first segment of the one or more segments that includes a static camera cut;
selecting, by the one or more processors, a first video frame of the first segment;
creating, by the one or more processors, a video frame canvas that is a second aspect ratio that is different than first aspect ratio, wherein the video frame canvas includes a first set of pixels including the first video frame and a second set of pixels provided adjacent to a boarder of the first video frame;
generating, by the one or more processors, a mask, the mask being a matrix including a first numerical value in a first set of positions in the matrix corresponding to a location of the first set of pixels and a second numerical value in a first set of positions in the matrix corresponding to a location of the second set of pixels;
causing, by the one or more processors and using a machine learning model, and based on the first video frame canvas, the first video frame to be modified into a second video frame, wherein the second set of pixels are modified in the second video frame to provide an appearance of the first video frame being of the second aspect ratio;
producing, by the one or more processors, a third video frame by overlaying the first video frame on the second video frame received from the machine learning model;
causing, by the one or more processors and using the machine learning model, the third video frame to be modified into a fourth video frame;
determining, by the one or more processors, that the machine learning model has generated an output video frame a threshold number of times; and
combining, based on the determination that the machine learning model has generated an output video frame a threshold number of times, output video frames generated by the machine learning model to produce second video content at the second aspect ratio.