US 11,758,082 B2
System for automatic video reframing
Lu Zhang, Dalian (CN); Jianming Zhang, Campbell, CA (US); Zhe Lin, Fremont, CA (US); and Radomir Meeh, Mountain View, CA (US)
Assigned to Adobe Inc., San Jose, CA (US)
Filed by Adobe Inc., San Jose, CA (US)
Filed on Nov. 15, 2021, as Appl. No. 17/526,853.
Application 17/526,853 is a continuation of application No. 16/900,435, filed on Jun. 12, 2020, granted, now 11,184,558.
Prior Publication US 2022/0078358 A1, Mar. 10, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. H04N 5/262 (2006.01); G11B 27/031 (2006.01); G06V 20/40 (2022.01); G06V 10/20 (2022.01); G06V 40/18 (2022.01)
CPC H04N 5/2628 (2013.01) [G06V 10/255 (2022.01); G06V 20/40 (2022.01); G06V 20/41 (2022.01); G11B 27/031 (2013.01); G06V 40/193 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
identifying, by a first neural network, a focal point within a mask of an object in a frame of a video segment;
receiving, from a second neural network, a reframing suggestion for the frame;
receiving a set of reframing parameters; and
reframing, by a reframing engine, each frame of the first video segment based on the focal point, the mask of the object, the reframing suggestion, and the set of reframing parameters.