A Performance Comparison of Loss Functions for Deep Face Recognition
This is an incremental study that benchmarks loss functions for face recognition, relevant for researchers and practitioners in computer vision.
The paper compares the performance of five loss functions (Cross-Entropy, Angular Softmax, Additive-Margin Softmax, ArcFace, and Marginal Loss) for deep face recognition using ResNet and MobileNet architectures on CASIA-Webface and MS-Celeb-1M datasets, with testing on LFW, but does not report specific numerical results.
Face recognition is one of the most widely publicized feature in the devices today and hence represents an important problem that should be studied with the utmost priority. As per the recent trends, the Convolutional Neural Network (CNN) based approaches are highly successful in many tasks of Computer Vision including face recognition. The loss function is used on the top of CNN to judge the goodness of any network. In this paper, we present a performance comparison of different loss functions such as Cross-Entropy, Angular Softmax, Additive-Margin Softmax, ArcFace and Marginal Loss for face recognition. The experiments are conducted with two CNN architectures namely, ResNet and MobileNet. Two widely used face datasets namely, CASIA-Webface and MS-Celeb-1M are used for the training and benchmark Labeled Faces in the Wild (LFW) face dataset is used for the testing.