CLFace: A Scalable and Resource-Efficient Continual Learning Framework for Lifelong Face Recognition

TL;DR

CLFace is a scalable, resource-efficient continual learning framework for face recognition that mitigates catastrophic forgetting and privacy issues by using feature-level distillation and a fixed model architecture.

Contribution

It introduces a novel continual learning approach for face recognition that eliminates the classification layer and employs multiple distillation schemes for incremental learning.

Findings

Outperforms baseline and state-of-the-art methods on benchmark datasets.

Effectively preserves knowledge of previous identities during incremental learning.

Maintains high recognition accuracy on unseen identities.

Abstract

An important aspect of deploying face recognition (FR) algorithms in real-world applications is their ability to learn new face identities from a continuous data stream. However, the online training of existing deep neural network-based FR algorithms, which are pre-trained offline on large-scale stationary datasets, encounter two major challenges: (I) catastrophic forgetting of previously learned identities, and (II) the need to store past data for complete retraining from scratch, leading to significant storage constraints and privacy concerns. In this paper, we introduce CLFace, a continual learning framework designed to preserve and incrementally extend the learned knowledge. CLFace eliminates the classification layer, resulting in a resource-efficient FR model that remains fixed throughout lifelong learning and provides label-free supervision to a student model, making it suitable for open-set face recognition during incremental steps. We introduce an objective function that employs feature-level distillation to reduce drift between feature maps of the student and teacher models across multiple stages. Additionally, it incorporates a geometry-preserving distillation scheme to maintain the orientation of the teacher model's feature embedding. Furthermore, a contrastive knowledge distillation is incorporated to continually enhance the discriminative power of the feature representation by matching similarities between new identities. Experiments on several benchmark FR datasets demonstrate that CLFace outperforms baseline approaches and state-of-the-art methods on unseen identities using both in-domain and out-of-domain datasets.