BoundaryFace: A mining framework with noise label self-correction for Face Recognition

TL;DR

BoundaryFace introduces a novel face recognition framework that incorporates noise label self-correction and hard sample mining, significantly improving performance on noisy large-scale datasets.

Contribution

It presents a new decision boundary-based mining framework with a noise label self-correction module, addressing label noise and hard sample mining in face recognition.

Findings

Outperforms state-of-the-art methods on multiple benchmarks.

Effectively handles datasets with high label noise.

Enhances discriminative feature learning through boundary-focused mining.

Abstract

Face recognition has made tremendous progress in recent years due to the advances in loss functions and the explosive growth in training sets size. A properly designed loss is seen as key to extract discriminative features for classification. Several margin-based losses have been proposed as alternatives of softmax loss in face recognition. However, two issues remain to consider: 1) They overlook the importance of hard sample mining for discriminative learning. 2) Label noise ubiquitously exists in large-scale datasets, which can seriously damage the model's performance. In this paper, starting from the perspective of decision boundary, we propose a novel mining framework that focuses on the relationship between a sample's ground truth class center and its nearest negative class center. Specifically, a closed-set noise label self-correction module is put forward, making this framework work well on datasets containing a lot of label noise. The proposed method consistently outperforms SOTA methods in various face recognition benchmarks. Training code has been released at https://github.com/SWJTU-3DVision/BoundaryFace.