Self-restrained Triplet Loss for Accurate Masked Face Recognition

TL;DR

This paper introduces a novel loss function and model to enhance masked face recognition accuracy, addressing challenges posed by face occlusion due to masks, especially relevant during the COVID-19 pandemic.

Contribution

The paper proposes the Self-restrained Triplet loss and Embedding Unmasking Model to improve masked face recognition performance on existing models and datasets.

Findings

Significant performance improvement on masked face datasets

Effective in both real and synthetic masked face scenarios

Enhances embeddings to resemble unmasked face representations

Abstract

Using the face as a biometric identity trait is motivated by the contactless nature of the capture process and the high accuracy of the recognition algorithms. After the current COVID-19 pandemic, wearing a face mask has been imposed in public places to keep the pandemic under control. However, face occlusion due to wearing a mask presents an emerging challenge for face recognition systems. In this paper, we present a solution to improve masked face recognition performance. Specifically, we propose the Embedding Unmasking Model (EUM) operated on top of existing face recognition models. We also propose a novel loss function, the Self-restrained Triplet (SRT), which enabled the EUM to produce embeddings similar to these of unmasked faces of the same identities. The achieved evaluation results on three face recognition models, two real masked datasets, and two synthetically generated masked face datasets proved that our proposed approach significantly improves the performance in most experimental settings.