US 12,142,028 B2
Object recognition apparatus and method based on environment matching
Ki-Min Yun, Daejeon (KR); Jin-Young Moon, Daejeon (KR); Jong-Won Choi, Seoul (KR); Joung-Su Youn, Seoul (KR); Seok-Jun Choi, Seoul (KR); and Woo-Seok Hyung, Seoul (KR)
Assigned to ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE, Daejeon (KR)
Filed by Electronics and Telecommunications Research Institute, Daejeon (KR)
Filed on Dec. 14, 2021, as Appl. No. 17/550,825.
Claims priority of application No. 10-2021-0092526 (KR), filed on Jul. 14, 2021.
Prior Publication US 2023/0015295 A1, Jan. 19, 2023
Int. Cl. G06V 10/764 (2022.01); G06V 10/762 (2022.01); G06V 20/40 (2022.01)
CPC G06V 10/764 (2022.01) [G06V 10/762 (2022.01); G06V 20/46 (2022.01); G06V 20/49 (2022.01)] 17 Claims
OG exemplary drawing
 
1. An object recognition apparatus based on environment matching, comprising:
a memory for storing at least one program; and
a processor for executing the program,
wherein the program performs:
acquiring a video that is input in real time to the object recognition apparatus;
determining a similarity between a key frame extracted from the video that is input in real time and each of training video data of a plurality of recognition models, based on a pretrained similarity-matching network;
selecting, among the plurality of recognition models, a recognition model trained with a video having a greatest similarity to the key frame extracted from the video that is input in real time; and
recognizing an object in a scene of the video that is input in real time using the selected recognition model as an initial model.