US 11,947,626 B2
Face recognition from unseen domains via learning of semantic features
Masoud Faraki, San Jose, CA (US); Xiang Yu, Mountain View, CA (US); Yi-Hsuan Tsai, Santa Clara, CA (US); Yumin Suh, Santa Clara, CA (US); and Manmohan Chandraker, Santa Clara, CA (US)
Assigned to NEC Corporation, Tokyo (JP)
Filed by NEC Laboratories America, Inc., Princeton, NJ (US)
Filed on Nov. 5, 2021, as Appl. No. 17/519,950.
Claims priority of provisional application 63/114,013, filed on Nov. 16, 2020.
Claims priority of provisional application 63/111,842, filed on Nov. 10, 2020.
Prior Publication US 2022/0147765 A1, May 12, 2022
Int. Cl. G06F 18/214 (2023.01); G06N 3/04 (2023.01); G06V 40/16 (2022.01)
CPC G06F 18/214 (2023.01) [G06N 3/04 (2013.01); G06V 40/161 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method for improving face recognition from unseen domains by learning semantically meaningful representations, the method comprising:
obtaining face images with associated identities from a plurality of datasets;
randomly selecting two datasets of the plurality of datasets to train a model;
sampling batch face images and their corresponding labels;
sampling triplet samples including one anchor face image, a sample face image from a same identity, and a sample face image from a different identity than that of the one anchor face image;
performing a forward pass by using the samples of the selected two datasets;
finding representations of the face images by using a backbone convolutional neural network (CNN);
generating covariances from the representations of the face images and the backbone CNN, the covariances made in different spaces by using positive pairs and negative pairs; and
employing the covariances to compute a cross-domain similarity loss function.