US 11,720,798 B2
Foreground-background-aware atrous multiscale network for disparity estimation
Xianzhi Du, San Jose, CA (US); Mostafa El-Khamy, San Diego, CA (US); and Jungwon Lee, San Diego, CA (US)
Assigned to Samsung Electronics Co., Ltd., Suwon-si (KR)
Filed by Samsung Electronics Co., Ltd., Suwon-si (KR)
Filed on May 27, 2021, as Appl. No. 17/332,792.
Application 17/332,792 is a continuation of application No. 16/451,524, filed on Jun. 25, 2019, granted, now 11,024,037.
Claims priority of provisional application 62/768,055, filed on Nov. 15, 2018.
Prior Publication US 2021/0287378 A1, Sep. 16, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 7/194 (2017.01); G06N 3/084 (2023.01); G06T 7/593 (2017.01); G06N 3/08 (2023.01); G06V 30/19 (2022.01); G06V 10/82 (2022.01); G06V 10/44 (2022.01)
CPC G06N 3/084 (2013.01) [G06N 3/08 (2013.01); G06T 7/194 (2017.01); G06T 7/596 (2017.01); G06V 10/454 (2022.01); G06V 10/82 (2022.01); G06V 30/1916 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A system for disparity estimation, the system comprising:
a feature extractor module configured to extract one or more feature maps from one or more input images;
a semantic information module configured to generate, based on one or more outputs of the feature extractor module, semantic information of a foreground object in the one or more input images;
a cost volume (CV) module configured to determine, based on one or more outputs of the feature extractor module, a cost volume; and
an atrous multiscale (AM) module configured to aggregate multiscale context information derived from one or more outputs of the cost volume module,
wherein the system is configured to estimate a disparity of the system based on the aggregated multiscale context information.