US 12,067,754 B2
Electronic device and method of inferring object in image
Kyounghoon Kim, Suwon-si (KR); Sihoon Song, Suwon-si (KR); and Sangbok Han, Suwon-si (KR)
Assigned to SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
Appl. No. 17/436,493
Filed by SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
PCT Filed Jul. 16, 2021, PCT No. PCT/KR2021/009201
§ 371(c)(1), (2) Date Sep. 3, 2021,
PCT Pub. No. WO2022/019580, PCT Pub. Date Jan. 27, 2022.
Claims priority of application No. 10-2020-0092544 (KR), filed on Jul. 24, 2020.
Prior Publication US 2023/0177794 A1, Jun. 8, 2023
Int. Cl. G06V 10/25 (2022.01); G06V 10/70 (2022.01); G06V 10/82 (2022.01)
CPC G06V 10/25 (2022.01) [G06V 10/82 (2022.01); G06V 10/87 (2022.01)] 16 Claims
OG exemplary drawing
 
1. A method of inferring an object in a moving image using a convolutional neural network (CNN) model, performed by an electronic device, the method comprising:
identifying a first region of interest in a first frame among a plurality of frames in the moving image, and a first object in the first region of interest, by providing the first frame to a plurality of convolution layer groups sequentially connected in the CNN model;
identifying a second region of interest in a second frame among the plurality of frames, the second region of interest corresponding to the first region of interest, and the second frame being after the first frame;
providing the second region of interest to the CNN model, and obtaining first output data that is output from a first convolution layer group from among the plurality of convolution layer groups; and
determining whether to identify a second object in the second region of interest by using a second convolution layer group from among the plurality of convolution layer groups, based on the first output data, wherein the second convolution layer group sequentially follows the first convolution layer group within the plurality of convolution layer groups.