Deep Learning for Face Anti-Spoofing: A Survey

Algorithm of face anti-spoofing based on pseudo-negative features generation

Introduction

Despite advancements in face anti-spoofing technology, attackers continue to pose challenges with their evolving deceptive methods. This is primarily due to the increased complexity of their attacks, coupled with a diversity in presentation modes, acquisition devices, and prosthetic materials. Furthermore, the scarcity of negative sample data exacerbates the situation by causing domain shift issues and impeding robust generalization. Hence, there is a pressing need for more effective cross-domain approaches to bolster the model's capability to generalize across different scenarios.

Methods

This method improves the effectiveness of face anti-spoofing systems by analyzing pseudo-negative sample features, expanding the training dataset, and boosting cross-domain generalization. By generating pseudo-negative features with a new algorithm and aligning these features with the use of KL divergence loss, we enrich the negative sample dataset, aiding the training of a more robust feature classifier and broadening the range of attacks that the system can defend against.

Results

Through experiments on four public datasets (MSU-MFSD, OULU-NPU, Replay-Attack, and CASIA-FASD), we assess the model's performance within and across datasets by controlling variables. Our method delivers positive results in multiple experiments, including those conducted on smaller datasets.

Discussion

Through controlled experiments, we demonstrate the effectiveness of our method. Furthermore, our approach consistently yields favorable results in both intra-dataset and cross-dataset evaluations, thereby highlighting its excellent generalization capabilities. The superior performance on small datasets further underscores our method's remarkable ability to handle unseen data beyond the training set.

Multi-Domain Feature Alignment for Face Anti-Spoofing

Face anti-spoofing is critical for enhancing the robustness of face recognition systems against presentation attacks. Existing methods predominantly rely on binary classification tasks. Recently, methods based on domain generalization have yielded promising results. However, due to distribution discrepancies between various domains, the differences in the feature space related to the domain considerably hinder the generalization of features from unfamiliar domains. In this work, we propose a multi-domain feature alignment framework (MADG) that addresses poor generalization when multiple source domains are distributed in the scattered feature space. Specifically, an adversarial learning process is designed to narrow the differences between domains, achieving the effect of aligning the features of multiple sources, thus resulting in multi-domain alignment. Moreover, to further improve the effectiveness of our proposed framework, we incorporate multi-directional triplet loss to achieve a higher degree of separation in the feature space between fake and real faces. To evaluate the performance of our method, we conducted extensive experiments on several public datasets. The results demonstrate that our proposed approach outperforms current state-of-the-art methods, thereby validating its effectiveness in face anti-spoofing.

PhysFormer++: Facial Video-Based Physiological Measurement with SlowFast Temporal Difference Transformer

Remote photoplethysmography (rPPG), which aims at measuring heart activities and physiological signals from facial video without any contact, has great potential in many applications (e.g., remote healthcare and affective computing). Recent deep learning approaches focus on mining subtle rPPG clues using convolutional neural networks with limited spatio-temporal receptive fields, which neglect the long-range spatio-temporal perception and interaction for rPPG modeling. In this paper, we propose two end-to-end video transformer based architectures, namely PhysFormer and PhysFormer++, to adaptively aggregate both local and global spatio-temporal features for rPPG representation enhancement. As key modules in PhysFormer, the temporal difference transformers first enhance the quasi-periodic rPPG features with temporal difference guided global attention, and then refine the local spatio-temporal representation against interference. To better exploit the temporal contextual and periodic rPPG clues, we also extend the PhysFormer to the two-pathway SlowFast based PhysFormer++ with temporal difference periodic and cross-attention transformers. Furthermore, we propose the label distribution learning and a curriculum learning inspired dynamic constraint in frequency domain, which provide elaborate supervisions for PhysFormer and PhysFormer++ and alleviate overfitting. Comprehensive experiments are performed on four benchmark datasets to show our superior performance on both intra- and cross-dataset testings. Unlike most transformer networks needed pretraining from large-scale datasets, the proposed PhysFormer family can be easily trained from scratch on rPPG datasets, which makes it promising as a novel transformer baseline for the rPPG community.