US 12,347,161 B2
System and method for dual-value attention and instance boundary aware regression in computer vision system
Qingfeng Liu, San Diego, CA (US); and Mostafa El-Khamy, San Diego, CA (US)
Assigned to Samsung Electronics Co., Ltd., Yongin-si (KR)
Filed by Samsung Electronics Co., Ltd., Suwon-si (KR)
Filed on Jul. 19, 2022, as Appl. No. 17/868,660.
Claims priority of provisional application 63/311,246, filed on Feb. 17, 2022.
Prior Publication US 2023/0260247 A1, Aug. 17, 2023
Int. Cl. G06K 9/00 (2022.01); G06T 7/11 (2017.01); G06V 10/26 (2022.01); G06V 10/52 (2022.01); G06V 10/77 (2022.01)
CPC G06V 10/52 (2022.01) [G06T 7/11 (2017.01); G06V 10/26 (2022.01); G06V 10/7715 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A computer vision system comprising:
one or more processors; and
memory comprising instructions that, when executed by the one or more processors, cause the one or more processors to:
determine a semantic multi-scale context feature and an instance multi-scale context feature of an input scene;
generate a joint attention map based on the semantic multi-scale context feature and the instance multi-scale context feature;
refine the semantic multi-scale context feature and instance multi-scale context feature based on the joint attention map; and
generate a panoptic segmentation image based on the refined semantic multi-scale context feature and the refined instance multi-scale context feature.