IntraLoss: Further Margin via Gradient-Enhancing Term for Deep Face Recognition

TL;DR

IntraLoss introduces a gradient-enhancing term to improve intra-class feature distribution in deep face recognition, leading to more compact and isotropic features and better margin between identities, outperforming state-of-the-art methods.

Contribution

The paper proposes IntraLoss, a novel gradient-enhancing method that explicitly improves intra-class feature distribution in face recognition models.

Findings

Outperforms state-of-the-art methods on LFW, YTF, and CFP-FP datasets.

Explicitly shrinks intra-class distribution, increasing inter-class margin.

Provides an intuitive geometric interpretation and easy integration with existing methods.

Abstract

Existing classification-based face recognition methods have achieved remarkable progress, introducing large margin into hypersphere manifold to learn discriminative facial representations. However, the feature distribution is ignored. Poor feature distribution will wipe out the performance improvement brought about by margin scheme. Recent studies focus on the unbalanced inter-class distribution and form a equidistributed feature representations by penalizing the angle between identity and its nearest neighbor. But the problem is more than that, we also found the anisotropy of intra-class distribution. In this paper, we propose the `gradient-enhancing term' that concentrates on the distribution characteristics within the class. This method, named IntraLoss, explicitly performs gradient enhancement in the anisotropic region so that the intra-class distribution continues to shrink, resulting in isotropic and more compact intra-class distribution and further margin between identities. The experimental results on LFW, YTF and CFP-FP show that our outperforms state-of-the-art methods by gradient enhancement, demonstrating the superiority of our method. In addition, our method has intuitive geometric interpretation and can be easily combined with existing methods to solve the previously ignored problems.