—Face Anti-spooﬁng (FAS) is essential to secure face recognition systems from various physical attacks. However, recent research generally focuses on short-distance applications ( i.e. , phone unlocking) while lacking consideration of long-distance scenes ( i.e. , surveillance security checks). In order to promote relevant research and ﬁll this gap in the community, we collect a large-scale Su rveillance Hi gh- Fi delity Mask (SuHi- FiMask) dataset captured under 40 surveillance scenes, which has 101 subjects from different age groups with 232 3D attacks (high- ﬁdelity masks), 200 2D attacks (posters, portraits, and screens), and 2 adversarial attacks. In this scene, low image resolution and noise interference are new challenges faced in surveillance FAS. Together with the SuHiFiMask dataset, we propose a Contrastive Quality-Invariance Learning (CQIL) network to alleviate the performance degradation caused by image quality from three aspects: (1) An Image Quality Variable module (IQV) is introduced to recover image information associated with discrimination by combining the super-resolution network. (2) Using generated sample pairs to simulate quality variance distributions to help contrastive learning strategies obtain robust feature representation under quality variation. (3) A Separate Quality Network (SQN) is designed to learn discriminative features independent of image quality. Finally, a large number of experiments verify the quality of the SuHiFiMask dataset and the superiority of the proposed CQIL.

Surveillance Face Anti-spoofing

Face anti-spoof systems are needed in facial recognition systems to ward off attacks that present fake faces in front of the camera or image capture sensor (presentation attack). To build the system, a data set is needed to build a classification model that distinguishes the authenticity of the face of the input image received by the system. In the past decade anti-face spoof research has produced many data sets that are public, but often researchers need time to build or use the right public data sets that are used to build facial anti-spoof models. This article conducts a literature study of public data sets using a systematic literature review method to find out the types of attacks that appear on the facial anti-spoof system, the development process, evolution, and availability of facial anti-spoof data sets. From the search and selection results based on the specified criteria, there were 42 primary research manuscripts in the period 2010 to 2021. The results of the literature study found that there were three trends in the development of anti-spoof facial data sets, namely, 1) data sets with a very large number, 2) datasets with different types of facial samples, and 3) datasets constructed with various devices and sensors. These various public data sets can be accessed freely but with special rules such as agreeing to an end user license agreement document from the researcher or the institution that owns the data set. However, there are also datasets that cannot be accessed due to invalid URLs or due to special rules from the cloud storage service provider where the datasets are stored.

Studi Literatur Presentation Attack dan Set Data Anti-Spoof Wajah

This Technical Briefing is published by The Alan Turing Institute’s Trustworthy Digital Infrastructure for Identity Systems project

Threat Models for Face Recognition Systems: Taxonomy of Threats

Through the release of three large-scale datasets and the successful holding of three competitions, we have promoted the development of the face anti-spoofing community. In this chapter, we will summarize our work in recent years from two aspects of datasets and competitions, including the characteristics of datasets CASIA-SURF, CASIA-SURF CeFA and CASIA-SURF HiFiMask, the preliminary solutions of benchmark methods MS-SEF, PSMM-Net, and CCL, and the core ideas of excellent competition algorithms. In addition, based on the development trend of competition algorithms in recent years and the demand for anti-spoofing technology in daily life, we determine the future work direction, including Flexible Modal Face Anti-Spoofing, Generalizable Face Anti-spoofing and Surveillance Face Anti-spoofing aspects. 

Conclusions and Future Work

Using multi-modal data, such as VIS, IR and Depth, for face anti-spoofing (FAS) is a robust and effective method, because complementary information between different modalities can better against various attacks. However, multi-modal data is difficult to obtain in application scenarios due to high costs, which makes the model trained with multi-modal data unavailable in the testing stage. We define this phenomenon as train-test inconsistency problem, which is ignored by most existing methods. To this end, we propose a novel multi-modal face anti-spoofing framework (GFF), which adopt multi-modal data during training, and only use a single modality during testing to simulate multi-modal input. Specifically, GFF is a two-step framework. In the step I, we adopt the GAN model to fit the face images distribution of different modalities, and learn the transform strategies between different distributions, so as to realize the generation from a single real modality image to other modalities. In the step II, we select the real face images in one modality and the generated images in the other modalities according to actual needs to construct a simulation dataset, which is used for training face anti-spoofing model. The advantage of GFF is that it has achieved a good trade-off between data capture cost and model performance in the real application scenarios. The experimental results show that the method proposed in this paper can effectively overcome the train-test inconsistency problem. On the CASIA-SRUF CeFA dataset, the performance of GFF surpasses the existing single-modality-based methods, and surprisingly surpasses some multi-modality-based methods.

Multi-modal Face Anti-spoofing Based on a Single Image

Face Anti-spoofing (FAS) is a challenging problem due to complex serving scenarios and diverse face presentation attack patterns. Especially when captured images are low-resolution, blurry, and coming from different domains, the performance of FAS will degrade significantly. The existing multi-modal FAS datasets rarely pay attention to the cross-domain problems under deployment scenarios, which is not conducive to the study of model performance. To solve these problems, we explore the fine-grained differences between multi-modal cameras and construct a cross-domain multi-modal FAS dataset under surveillance scenarios called GREAT-FASD-S. Besides, we propose an Attention based Face Anti-spoofing network with Feature Augment (AFA) to solve the FAS towards low-quality face images. It consists of the depthwise separable attention module (DAM) and the multi-modal based feature augment module (MFAM). Our model can achieve state-of-the-art performance on the CASIA-SURF dataset and our proposed GREAT-FASD-S dataset.

A Dataset and Benchmark Towards Multi-Modal Face Anti-Spoofing Under Surveillance Scenarios

Face Anti-spoofing (FAS) is a challenging problem due to the complex serving scenario and diverse face presentation attack patterns. Using single modal images which are usually captured with RGB cameras is not able to deal with the former because of serious overfitting problems. The existing multi-modal FAS datasets rarely pay attention to the cross domain problems, training FAS system on these data leads to inconsistencies and low generalization capabilities in deployment since imaging princi-ples(structured light, TOF, etc.) and pre-processing methods vary between devices. We explore the subtle fine-grained differences betweeen multi-modal cameras and proposed a cross domain multi-modal FAS dataset GREAT-FASD and several evaluation protocols for academic community. Furthermore, we incorporate the multiplicative attention and center loss to enhance the representative power of CNN via seeking out complementary information as a powerful baseline. In addition, extensive experiments have been conducted on the proposed dataset to analyze the robustness to distinguish spoof faces and bona-fide faces. Experimental results show the effectiveness of proposed method and achieve the state-of-the-art competitive results. Finally, we visualize our future distribution in hidden space andobserve that the proposed method is able to lead the network to generate a large margin for face anti-spoofing task. Index Terms-Face Anti-spoofing, Multi-modal, Cross Domain, Convolutional Neural Network

Qiaobin Ji

Papers

A Dataset and Benchmark Towards Multi-Modal Face Anti-Spoofing Under Surveillance Scenarios

A Cross Domain Multi-modal Dataset for Robust Face Anti-spoofing