Rehearsal-Free Domain Continual Face Anti-Spoofing: Generalize More and Forget Less

TL;DR

This paper introduces a novel rehearsal-free domain continual face anti-spoofing method that enhances generalization to unseen domains and reduces forgetting of previous knowledge without needing past data, using innovative adapters and regularization techniques.

Contribution

It presents the first rehearsal-free approach for domain continual face anti-spoofing, combining a dynamic adapter and proxy prototype regularization for better generalization and forgetting mitigation.

Findings

Improves generalization to unseen domains.

Reduces catastrophic forgetting without previous data.

Outperforms existing methods in continual learning scenarios.

Abstract

Face Anti-Spoofing (FAS) is recently studied under the continual learning setting, where the FAS models are expected to evolve after encountering the data from new domains. However, existing methods need extra replay buffers to store previous data for rehearsal, which becomes infeasible when previous data is unavailable because of privacy issues. In this paper, we propose the first rehearsal-free method for Domain Continual Learning (DCL) of FAS, which deals with catastrophic forgetting and unseen domain generalization problems simultaneously. For better generalization to unseen domains, we design the Dynamic Central Difference Convolutional Adapter (DCDCA) to adapt Vision Transformer (ViT) models during the continual learning sessions. To alleviate the forgetting of previous domains without using previous data, we propose the Proxy Prototype Contrastive Regularization (PPCR) to constrain the continual learning with previous domain knowledge from the proxy prototypes. Simulate practical DCL scenarios, we devise two new protocols which evaluate both generalization and anti-forgetting performance. Extensive experimental results show that our proposed method can improve the generalization performance in unseen domains and alleviate the catastrophic forgetting of the previous knowledge. The codes and protocols will be released soon.