Custom silicone Face Masks: Vulnerability of Commercial Face Recognition Systems &amp; Presentation Attack Detection

Ramachandra, Raghavendra; Venkatesh, Sushma; Raja, Kiran B.; Bhattacharjee, Sushil; Wasnik, Pankaj; Marcel, Sébastien; Busch, Christoph

doi:10.1109/iwbf.2019.8739236

Cited by 27 publications

(18 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For this study, Nimba Creations Ltd., a special-effects company, has constructed very realistic silicone masks for 17 target-subjects, based on pictures and 3D scans provided by us. State of the art CNN-FR systems have been shown to be highly vulnerable to PAs made using such masks [1,2]. In this work we present a new method for detecting impersonation PAs made using custom silicone masks.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Multispectral Deep Embeddings as a Countermeasure to Custom Silicone Mask Presentation Attacks

Kotwal

Bhattacharjee

Marcel

2019

IEEE Trans. Biom. Behav. Identity Sci.

Self Cite

View full text Add to dashboard Cite

This work focuses on detecting presentation attacks (PA) mounted using custom silicone masks. Face recognition (FR) systems have been shown to be highly vulnerable to PAs based on such masks [1, 2]. Here we explore the use of multispectral data (color imagery, near infrared (NIR) imagery and thermal imagery) for face presentation attack detection (PAD), specifically against the custom silicone mask attacks. Using a new dataset (XCSMAD) representing 21 custom made masks, we establish the baseline performance of several commonly used face-PAD methods, on the different imaging channels. Considering thermal imagery in particular, our experiments show that low-cost thermal imaging devices are as effective in face-PAD as more expensive thermal cameras, for mask-based attacks. This result reinforces the case for the use of thermal data in face-PAD. We also demonstrate that fusing information from multiple channels leads to significant improvement in face-PAD performance. Finally, we propose a new approach to face-PAD of custom silicone masks using a convolutional neural network (CNN). On individual spectral channels, the proposed approach achieves state-of-the-art results. Using multispectralfusion, the proposed CNN-based method significantly outperforms the baseline methods. The new dataset and source-code for our experiments is freely available for research purposes.

show abstract

Section: Resultsmentioning

confidence: 99%

“…Previous studies have shown that FR systems are highly vulnerable to customsilicone-mask based PAs [1,2]. One underlying reason for this vulnerability is that these FR systems are designed to operate on visible-light (RGB-color) imagery.…”

Section: Introductionmentioning

confidence: 99%

Multispectral Deep Embeddings as a Countermeasure to Custom Silicone Mask Presentation Attacks

Kotwal

Bhattacharjee

Marcel

2019

IEEE Trans. Biom. Behav. Identity Sci.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Ramachandra et al [8] presented An observational research on vulnerability identification and appearance attack detection employing custom 3D silicone masks leading to actual targets for commercial face recognition systems (FRS). Mahore et al [9] illustrated a modern method for detecting the appearance of an anti-spoofing-based 3D face mask by using intensity and texture attribute descriptors.…”

Section: Related Workmentioning

confidence: 99%

“…Deep learning algorithms (DL) allowed the development of highly precise systems and became a standard option for the analysis of different complex applications [2], [3] in different areas. DL methods have been extremely effective over the last decade and have been widely used to build artificial intelligence [6][7][8] in nearly every domain. Deep convolutional neural networks have recently been successfully applied to face masks detection.…”

Section: Introductionmentioning

confidence: 99%

Face Mask Detection Classifier and Model Pruning with Keras-Surgeon

Negi

Chauhan

Kumar

et al. 2020

2020 5th IEEE International Conference on Recent Advances and Innovations in Engineering (ICRAIE)

View full text Add to dashboard Cite

Multidisciplinary initiatives in the new world of coronavirus were combined to limit the spread of the pandemic. Interestingly, the AI group was a part of those efforts. This result-based approach is used to help scan, assess, predict and track current patients and possibly potential patients. Developments for tracking social distances or recognizing face masks have made headlines in particular. Most current advanced approaches to face mask recognition are built based on deep learning which is dependent on a large number of face samples. Nearly everybody wears a mask during corona virus outbreak in order to effectively avoid the spread of COVID-19 virus. Our goal is to train a customized deep learning model that helps to detect even if or not a person wears a mask and study the concept of model pruning with Keras-Surgeon. Model pruning can be efficient in reducing model size, so that it can be easily implemented and inferred on embedded systems.

show abstract

“…With the appearance of custom made face masks and deepfake videos trained by generative adversarial network [11], PAD meets new challenges. A recent study shows that CNN based face recognition systems are vulnerable to custom silicone masks [12] and several PAD techniques are validated on custom masks database [13].…”

Section: Introductionmentioning

confidence: 99%

An Auxiliary Method Based on Hyperspectral Reflectance for Presentation Attack Detection

Li¹,

Ardabilian²,

Zine³

2019

2019 15th International Conference on Signal-Image Technology &Amp; Internet-Based Systems (SITIS)

View full text Add to dashboard Cite

Face recognition has reached a high accuracy in recent years by adopting convolutional neural networks. However, it suffers from presentation attacks such as 2D face photos, and 3D masks. The vulnerability of face recognition and presentation attacks detection (PAD) attract numerous researchers in recent years. Most studies have only focused on PAD algorithms by analyzing texture information, depth information or thermal images. On the other hand, hyperspectral reflectance, which benefits from the development of line-scan HSI sensors, makes it possible to detect information about the inner structure of materials. Our research proposes an auxiliary method to support face recognition by analyzing hyperspectral reflectance. Combined with biological facts of human skin, we trained a neural network with pigmentation fractions inside human skin and corresponding reflectance. The results show high accuracy in identifying skin and non-skin.

show abstract

Custom silicone Face Masks: Vulnerability of Commercial Face Recognition Systems & Presentation Attack Detection

Cited by 27 publications

References 9 publications

Multispectral Deep Embeddings as a Countermeasure to Custom Silicone Mask Presentation Attacks

Multispectral Deep Embeddings as a Countermeasure to Custom Silicone Mask Presentation Attacks

Face Mask Detection Classifier and Model Pruning with Keras-Surgeon

An Auxiliary Method Based on Hyperspectral Reflectance for Presentation Attack Detection

Contact Info

Product

Resources

About