Representation constraint-based dual-channel network for face antispoofing

Although multimodal face data have obvious advantages in describing live and spoofed features, single-modality face antispoofing technologies are still widely used when it is difficult to obtain multimodal face images or inconvenient to integrate and deploy multimodal sensors. Since the live/spoofed representations in visible light facial images include considerable face identity information interference, existing deep learning-based face antispoofing models achieve poor performance when only the visible light modality is used. To address the above problems, the authors design a dual-channel network structure and a constrained representation learning method for face antispoofing. First, they design a dual-channel attention mechanism-based grouped convolutional neural network (CNN) to learn important deceptive cues in live and spoofed faces. Second, they design inner contrastive estimation-based representation constraints for both live and spoofed samples to minimise the sample similarity loss to prevent the CNN from learning more facial appearance information. This increases the distance between live and spoofed faces and enhances the network's ability to identify deceptive cues. The evaluation results indicate that the framework we designed achieves an average classification error rate (ACER) of 2.37% on the visible light modality subset of the CASIA-SURF dataset and an ACER of 2.4% on the CASIA-SURF CeFA dataset, outperforming existing methods. The proposed method achieves low ACER scores in cross-dataset testing, demonstrating its advantage in domain generalisation. A grouped convolutional neural network based on the dual-channel attention mechanism is proposed to learn the deception cues in single-modality live and spoofed faces. An ICE representation constraint is designed to minimise the distribution distance of similar samples, eliminate the interference of facial appearance information, and improve the face antispoofing performance.image

Automated summary

This study tackles the problem of poor performance of single-modality face antispoofing models when only visible light is used. The authors design a dual-channel network structure and a constrained representation learning method to address this issue. By using a dual-channel attention mechanism-based grouped CNN, they learn deceptive cues in both live and spoofed faces, and design an inner contrastive estimation-based representation constraint to enhance the network's ability to identify deceptive cues. The evaluation results show that the proposed method outperforms existing methods with low error rates in cross-dataset testing, demonstrating its advantage in domain generalisation.