Aurora Guard: Real-Time Face Anti-Spoofing via Light Reflection

Liu, Yao; Tai, Ying; Li, Jilin; Ding, Shouhong; Wang, Chengjie; Huang, Feiyue; Li, Dongyang; Qi, Wenshuai; Ji, Rongrong

doi:10.48550/arxiv.1902.10311

Cited by 4 publications

(5 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our dataset outperforms previous public datasets [27], [28], [34] in three points: 1) Our dataset is the largest one that includes 30,000 live and spoofing videos (average duration to be 2s), collected from 10000 subjects, compared to 12,000 videos from 200 subjects in [18]. 2) As shown in Fig.…”

Section: Dataset Collectionmentioning

confidence: 87%

“…[14]- [17] achieved better effects by considering the inter-frame information, whereas the feature of face abnormal clues is not be utilized. [18] label face depth with the 3D camera equipped to focus on face detail, which is difficult to promote. By contrast, [19]- [21] convert 2D faces to 3D faces using PRNet, which has a lower label cost but a higher error rate.…”

Section: Introductionmentioning

confidence: 99%

“…We propose a feature-compressed attention module (FCAM) to process features in the temporal dimension. To balance the cost and accuracy, we propose a lightweight face labeling method based on face landmarks, which is faster than [18] and more accurate than [19]- [21]. The face position map obtained by the proposed labeling method is easier to model at fine-grain than depth map.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Multi-Frames Temporal Abnormal Clues Learning Method for Face Anti-Spoofing

Cong¹,

Zhang²,

He³

et al. 2022

Proceedings of the 34th International Conference on Software Engineering and Knowledge Engineering

View full text Add to dashboard Cite

Face anti-spoofing researches are widely used in face recognition and has received more attention from industry and academics. In this paper, we propose the EulerNet, a new temporal feature fusion network in which the differential filter and residual pyramid are used to extract and amplify abnormal clues from continuous frames, respectively. A lightweight sample labeling method based on face landmarks is designed to label large-scale samples at a lower cost and has better results than other methods such as 3D camera. Finally, we collect 30,000 live and spoofing samples using various mobile ends to create a dataset that replicates various forms of attacks in a real-world setting. Extensive experiments on public OULU-NPU show that our algorithm is superior to the state of art and our solution has already been deployed in real-world systems servicing millions of users.

show abstract

Section: Dataset Collectionmentioning

confidence: 87%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Multi-Frames Temporal Abnormal Clues Learning Method for Face Anti-Spoofing

Cong¹,

Zhang²,

He³

et al. 2022

Proceedings of the 34th International Conference on Software Engineering and Knowledge Engineering

View full text Add to dashboard Cite

show abstract

“…Resnet features are extracted from each modalities output and classified using softmax classifier to real or fake face. The method is able to detect photo spoof attack but fails for video replay attacks.Liu et al [20] proposed a face anti-spoof method using light reflection properties. A random sequence of light cues and intensities called as light captcha is generated and screen is manipulated to cast light as mentioned in light captcha.…”

Section: Related Workmentioning

confidence: 99%

Deep Learning Based Challenge Response Liveliness Matching for Presentation Attack Detection in Face Recognition Biometric Authentication Systems

R¹,

K²,

Y³

2022

INDJCSE

View full text Add to dashboard Cite

Face recognition based biometric authentication systems are being widely adopted but they are vulnerable to presentation attacks. Detecting presentation attack is important to enhance the security level of face recognition biometric systems. Many presentation attack detection systems (PAD) have been proposed based on comparison of real and presentation image features. But these solutions can be deceived easily by creating the exact replica of real face. To solve this problem, this work proposes a liveliness approach which solves PAD as a challenge response problem. The response of face to a challenge is measured and analyzed to detect PAD. The challenge response matching is realized using a novel Face action unit biased convolutional neural network which selectively skips feature learning in non action unit areas. This novel deep learning model speeds up the challenge response face matching, increases the accuracy of liveliness matching and robust against environmental distortionss.

show abstract

“…Liu et al [69] likewise proposed a method based on light reflection, named Aurora Guard (AG). This method utilizes light reflection to impose two auxiliary information: light parameter sequence and depth map.…”

Section: Face Reflectancementioning

confidence: 99%

Deep Learning meets Liveness Detection: Recent Advancements and Challenges

Sabaghi¹,

Oghbaie²,

Kooshan³

et al. 2021

Preprint

View full text Add to dashboard Cite

Facial biometrics has been recently received tremendous attention as a convenient replacement for traditional authentication systems. Consequently, detecting malicious attempts has found great significance, leading to extensive studies in face anti-spoofing (FAS),i.e., face presentation attack detection. Deep feature learning and techniques, as opposed to hand-crafted features, have promised dramatic increase in the FAS systems' accuracy, tackling the key challenges of materializing realworld application of such systems. Hence, a new research area dealing with development of more generalized as well as accurate models is increasingly attracting the attention of the research community and industry. In this paper, we present a comprehensive survey on the literature related to deep-feature-based FAS methods since 2017. To shed light on this topic, a semantic taxonomy based on various features and learning methodologies is represented. Further, we cover predominant public datasets for FAS in a chronological order, their evolutional progress, and the evaluation criteria (both intra-dataset and inter-dataset). Finally, we discuss the open research challenges and future directions.

show abstract

Aurora Guard: Real-Time Face Anti-Spoofing via Light Reflection

Cited by 4 publications

References 18 publications

Multi-Frames Temporal Abnormal Clues Learning Method for Face Anti-Spoofing

Multi-Frames Temporal Abnormal Clues Learning Method for Face Anti-Spoofing

Deep Learning Based Challenge Response Liveliness Matching for Presentation Attack Detection in Face Recognition Biometric Authentication Systems

Deep Learning meets Liveness Detection: Recent Advancements and Challenges

Contact Info

Product

Resources

About