US 12,223,690 B2
Face clustering method and apparatus, classification storage method, medium and electronic device
Wenhao Zhang, Beijing (CN); Hanwen Liu, Beijing (CN); and Jingtao Xu, Beijing (CN)
Assigned to BOE TECHNOLOGY GROUP CO., LTD., Beijing (CN)
Appl. No. 17/773,123
Filed by BOE TECHNOLOGY GROUP CO., LTD., Beijing (CN)
PCT Filed Mar. 18, 2021, PCT No. PCT/CN2021/081539
§ 371(c)(1), (2) Date Apr. 29, 2022,
PCT Pub. No. WO2022/193232, PCT Pub. Date Sep. 22, 2022.
Prior Publication US 2023/0245421 A1, Aug. 3, 2023
Int. Cl. G06V 10/762 (2022.01); G06V 10/74 (2022.01); G06V 10/776 (2022.01); G06V 10/98 (2022.01); G06V 40/16 (2022.01)
CPC G06V 10/763 (2022.01) [G06V 10/761 (2022.01); G06V 10/776 (2022.01); G06V 10/98 (2022.01); G06V 40/172 (2022.01)] 19 Claims
OG exemplary drawing
 
1. A face clustering method, comprising acquiring and clustering a face image to be clustered, the clustering comprising:
acquiring a similarity threshold corresponding to a quantity level of current image categories in an image category library, wherein at least two different quantity levels correspond to different similarity thresholds;
acquiring a similarity between the face image to be clustered and an image of at least one image category in the image category library;
judging, according to a current similarity threshold and the similarity, whether there is an image of the same category as the face image to be clustered in the image category library; determining, when there is the image of the same category as the face image to be clustered in the image category library, a category label of the face image to be clustered according to a category label of the image of the same category as the face image to be clustered; and assigning, when there is no image of the same category as the face image to be clustered in the image category library, the category label to the face image to be clustered according to a first preset rule,
wherein the acquiring the similarity threshold corresponding to the quantity level of current image categories in the image category library comprises:
judging whether the quantity level of image categories in the image category library is changed; adopting, when the quantity level of image categories in the image category library is not changed, a similarity threshold used when a previous face image to be clustered is clustered; and updating, when the quantity level of image categories in the image category library is changed, the similarity threshold according to the changed quantity level.