CosFace: Large Margin Cosine Loss for Deep Face Recognition

TL;DR

This paper introduces CosFace, a novel large margin cosine loss function that enhances face recognition accuracy by improving feature discrimination through normalization and angular margin maximization, achieving state-of-the-art results.

Contribution

The paper proposes a new loss function, LMCL, reformulating softmax as a cosine loss with a margin, leading to better discrimination in face recognition tasks.

Findings

Achieves state-of-the-art performance on multiple face recognition benchmarks.

Effectively maximizes inter-class variance and minimizes intra-class variance.

Demonstrates the superiority of cosine margin loss over previous methods.

Abstract

Face recognition has made extraordinary progress owing to the advancement of deep convolutional neural networks (CNNs). The central task of face recognition, including face verification and identification, involves face feature discrimination. However, the traditional softmax loss of deep CNNs usually lacks the power of discrimination. To address this problem, recently several loss functions such as center loss, large margin softmax loss, and angular softmax loss have been proposed. All these improved losses share the same idea: maximizing inter-class variance and minimizing intra-class variance. In this paper, we propose a novel loss function, namely large margin cosine loss (LMCL), to realize this idea from a different perspective. More specifically, we reformulate the softmax loss as a cosine loss by $L_2$ normalizing both features and weight vectors to remove radial variations, based on which a cosine margin term is introduced to further maximize the decision margin in the angular space. As a result, minimum intra-class variance and maximum inter-class variance are achieved by virtue of normalization and cosine decision margin maximization. We refer to our model trained with LMCL as CosFace. Extensive experimental evaluations are conducted on the most popular public-domain face recognition datasets such as MegaFace Challenge, Youtube Faces (YTF) and Labeled Face in the Wild (LFW). We achieve the state-of-the-art performance on these benchmarks, which confirms the effectiveness of our proposed approach.