Multi-Modal Face Presentation Attack Detection via Spatial and Channel Attentions

Wang, Guoqing; Lan, Chuanxin; Han, Hu; Shan, Shiguang; Chen, Xilin

doi:10.1109/cvprw.2019.00200

Cited by 35 publications

(13 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There is a pooling layer (max pooling or CDP) with or without spatial attention [73] after each cell. The attention module forces the cells to learn more concentrated features, which is proved to be effective [74] for FAS task. Finally, the low-mid-high level features are concatenated for prediction.…”

Section: Fas Search Spacementioning

confidence: 99%

NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing

Wan

Qin

et al. 2021

IEEE Trans. Pattern Anal. Mach. Intell.

165

102

View full text Add to dashboard Cite

Face anti-spoofing (FAS) plays a vital role in securing face recognition systems. Existing methods heavily rely on the expert-designed networks, which may lead to a sub-optimal solution for FAS task. Here we propose the first FAS method based on neural architecture search (NAS), called NAS-FAS, to discover the well-suited task-aware networks. Unlike previous NAS works mainly focus on developing efficient search strategies in generic object classification, we pay more attention to study the search spaces for FAS task. The challenges of utilizing NAS for FAS are in two folds: the networks searched on 1) a specific acquisition condition might perform poorly in unseen conditions, and 2) particular spoofing attacks might generalize badly for unseen attacks. To overcome these two issues, we develop a novel search space consisting of central difference convolution and pooling operators. Moreover, an efficient static-dynamic representation is exploited for fully mining the FAS-aware spatio-temporal discrepancy. Besides, we propose Domain/Type-aware Meta-NAS, which leverages cross-domain/type knowledge for robust searching. Finally, in order to evaluate the NAS transferability for cross datasets and unknown attack types, we release a large-scale 3D mask dataset, namely CASIA-SURF 3DMask, for supporting the new 'cross-dataset cross-type' testing protocol. Experiments demonstrate that the proposed NAS-FAS achieves state-of-the-art performance on nine FAS benchmark datasets with four testing protocols.

show abstract

Section: Fas Search Spacementioning

confidence: 99%

NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing

Wan

Qin

et al. 2021

IEEE Trans. Pattern Anal. Mach. Intell.

165

102

View full text Add to dashboard Cite

show abstract

“…Cai et al [10] discriminating local features for face spoofing detection Works only for artificial faces Yu et al [11] used material characteristics detected from reflection to classify spoofed images Accuracy disrupted by reflecting noise along with image Tu et al [12] Used motion cues in eye, mouth and head to detect spoofing Fails in presence of video replays Wang et al [13] acquired face image in four different modalities and detected liveliness Fails for video replay attacks Liu et al [14] used sequence of random light intensities and their reflection properties…”

Section: A Liveliness Detectionmentioning

confidence: 99%

Challenge Responsive Multi Modal Biometric Authentication Resilient to Presentation Attacks

2022

IJIES

View full text Add to dashboard Cite

Multimodal biometric systems are a natural evolution of traditional biometric authentication systems with use of multiple biometric to enhance the security level. Presentation attacks in form of photos and replay videos are becoming a threat to the biometric authentication systems and it becomes necessary to prevent from these attacks for an enhanced security level. This work proposes a challenge response based multi modal biometric authentication resilient to presentation attack. The proposed system is designed for multi modal biometric involving face and iris biometrics with challenge in form of arbitrary emotion invoking image at a random position on the screen. The changes in region of interest around the Facial landmarks and distance of iris to Facial landmarks are captured as response. The response is then matched with expected response derived based on the position of arbitrary emotion invoking image on screen to detect presentation attacks. In addition the work also proposes a hybrid deep feature which provides an addition security in recognition process to ensure authentication failure for fake samples. The error rate in leakage of spoofs is very low in proposed solution (less than 1%) in different environments compared to existing works. The recognition accuracy of face is atleast 1% higher and atleast 2% higher for IRIS in proposed solution compared to existing works.

show abstract

“…Compared with the previous competitions [2,6,9], the majority of the final participants (10 out of 13) of this competition came from the industry, which indicates the increased need for realiable liveness detection products in daily life applications. Furthermore, one highlight of the CVPR2019 challenge is that the three top-performing teams (VisionLabs 3 , ReadSense 4 , and Feather 5 ) released their source code in GitHub and summarized their approaches in the related CVPR workshop papers [55,63,70,87], enhancing the fairness, transparency, and reproducibility of the solutions so that they can be easily facilitated by the face recognition community. Fig.…”

Section: Multi-modal Face Anti-spoofing Attack Detection Challenge (C...mentioning

confidence: 99%

Review of Face Presentation Attack Detection Competitions

Yu¹,

Komulainen²,

Li³

et al. 2021

Preprint

View full text Add to dashboard Cite

Face presentation attack detection (PAD) has received increasing attention ever since the vulnerabilities to spoofing have been widely recognized. The state of the art in unimodal and multi-modal face anti-spoofing has been assessed in eight international competitions organized in conjunction with major biometrics and computer vision conferences in 2011, 2013, 2017, 2019, 2020 and 2021, each introducing new challenges to the research community. In this chapter, we present the design and results of the five latest competitions from 2019 until 2021. The first two challenges aimed to evaluate the effectiveness of face PAD in multi-modal setup introducing near-infrared (NIR) and depth modalities in addition to colour camera data, while the latest three competitions focused on evaluating domain and attack type generalization abilities of face PAD algorithms operating on conventional colour images and videos. We also discuss the lessons learnt from the competitions and future challenges in the field in general.

show abstract

Multi-Modal Face Presentation Attack Detection via Spatial and Channel Attentions

Cited by 35 publications

References 25 publications

NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing

NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing

Challenge Responsive Multi Modal Biometric Authentication Resilient to Presentation Attacks

Review of Face Presentation Attack Detection Competitions

Contact Info

Product

Resources

About