US 11,972,578 B2
Method and system for object tracking using online training
Myunggu Kang, Seongnam-si (KR); Dongyoon Wee, Seongnam-si (KR); and Soonmin Bae, Seongnam-si (KR)
Assigned to NAVER CORPORATION, Seongnam-si (KR)
Filed by NAVER CORPORATION, Seongnam-si (KR)
Filed on Aug. 27, 2021, as Appl. No. 17/458,896.
Application 17/458,896 is a continuation of application No. PCT/KR2020/001866, filed on Feb. 11, 2020.
Claims priority of application No. 10-2019-0023916 (KR), filed on Feb. 28, 2019.
Prior Publication US 2021/0390347 A1, Dec. 16, 2021
Int. Cl. G06T 7/246 (2017.01); G06F 18/214 (2023.01); G06F 18/22 (2023.01); G06F 18/2413 (2023.01); G06N 20/00 (2019.01); G06T 7/73 (2017.01); G06V 10/62 (2022.01); G06V 20/40 (2022.01)
CPC G06T 7/251 (2017.01) [G06F 18/2148 (2023.01); G06T 7/246 (2017.01); G06T 7/73 (2017.01); G06V 10/62 (2022.01); G06V 20/48 (2022.01)] 9 Claims
OG exemplary drawing
 
1. An object tracking method for tracking an object in video performed by at least one processor configured to execute computer-readable instructions stored in a memory, the object tracking method comprising:
training a classifier model using global pattern matching; and
classifying and tracking each target through online training including the classifier model;
wherein the training of the classifier model comprises,
separating a valid period in which all the targets are present in an entire consecutive period of an input video;
generating first training data after labelling a single valid period among the valid periods and training the classifier model; and
generating second training data after labelling a next valid period, generating accumulated training data by merging the first training data and the second training data, and repeatedly training the classifier model.