This paper addresses the problem of face presentation attack detection using different image modalities. In particular, the usage of short wave infrared (SWIR) imaging is considered. Face presentation attack detection is performed using recent models based on Convolutional Neural Networks using only carefully selected SWIR image differences as input. Conducted experiments show superior performance over similar models acting on either color images or on a combination of different modalities (visible, NIR, thermal and depth), as well as on a SVM-based classifier acting on SWIR image differences. Experiments have been carried on a new public and freely available database, containing a wide variety of attacks. Video sequences have been recorded thanks to several sensors resulting in 14 different streams in the visible, NIR, SWIR and thermal spectra, as well as depth data. The best proposed approach is able to almost perfectly detect all impersonation attacks while ensuring low bonafide classification errors. On the other hand, obtained results show that obfuscation attacks are more difficult to detect. We hope that the proposed database will foster research on this challenging problem. Finally, all the code and instructions to reproduce presented experiments is made available to the research community.

Deep Models and Shortwave Infrared Information to Detect Face Presentation Attacks

Face recognition algorithms have demonstrated very high recognition performance, suggesting suitability for real world applications Despite the enhanced accuracies, robustness of these algorithms against attacks and bias has been challenged This paper summarizes different ways in which the robustness of a face recognition algorithm is challenged, which can severely affect its intended working Different types of attacks such as physical presentation attacks, disguise/makeup, digital adversarial attacks, and morphing/tampering using GANs have been discussed We also present a discussion on the effect of bias on face recognition models and showcase that factors such as age and gender variations affect the performance of modern algorithms The paper also presents the potential reasons for these challenges and some of the future research directions for increasing the robustness of face recognition models

/pdf/on-the-robustness-of-face-recognition-algorithms-against-4pm6ra3h9o.pdf

On the Robustness of Face Recognition Algorithms Against Attacks and Bias

Face recognition is diversely used in modern biometric and security applications. Most of the current face recognition techniques show good results in a constrained environment. However, these techniques face many problems in real-world scenarios such as low-quality images, temporal variations and facial disguises creating variations in facial features. The reason for these deteriorating results is the employment of handcrafted features having weak generalization capabilities and neglecting the complexities associated with domain adaption in case of deep learning models. In this paper, we have studied the efficacy of deep learning methods incorporating simple noise-based data augmentation for disguise invariant face recognition (DIFR). The proposed method detects face in an image using Viola Jones face detector and classifies it using a pre-trained Convolutional Neural Network (CNN) fine-tuned for DIFR. During transfer learning, a pre-trained CNN learns generalized disguise-invariant features from facial images of several subjects to correctly identify them under varying facial disguises. We have compared four different pre-trained 2D CNNs, each with different number of learning parameters, based on their classification accuracy and execution time for selecting a suitable model for DIFR. Comprehensive experiments and comparative analysis have been conducted on six challenging facial disguise datasets. Resnet-18 gives the best trade-off between accuracy and efficiency, by achieving an average accuracy of 98.19% with an average execution time of 0.32 seconds. The promising results achieved in these experiments reflect the efficiency of the proposed method and outperforms the existing methods in all aspects.

An automated and efficient convolutional architecture for disguise-invariant face recognition using noise-based data augmentation and deep transfer learning

In the last decades, the broad development experienced by biometric systems has unveiled several threats which may decrease their trustworthiness. Those are attack presentations which can be easily carried out by a non-authorised subject to gain access to the biometric system. In order to mitigate those security concerns, most face Presentation Attack Detection techniques have reported a good detection performance when they are evaluated on known Presentation Attack Instruments (PAI) and acquisition conditions, in contrast to more challenging scenarios where unknown attacks are included in the test set. For those more realistic scenarios, the existing algorithms face difficulties to detect unknown PAI species in many cases. In this work, we use a new feature space based on Fisher Vectors, computed from compact Binarised Statistical Image Features histograms, which allow discovering semantic feature subsets from known samples in order to enhance the detection of unknown attacks. This new representation, evaluated for challenging unknown attacks taken from freely available facial databases, shows promising results: a BPCER100 under 17% together with an AUC over 98% can be achieved in the presence of unknown attacks. In addition, by training a limited number of parameters, our method is able to achieve state-of-the-art deep learning-based approaches for cross-dataset scenarios.

On the generalisation capabilities of Fisher vector-based face presentation attack detection

In multi-view 3D object retrieval tasks, it is pivotal to aggregate visual features extracted from multiple view images to generate a discriminative representation for a 3D object. The existing multi-view convolutional neural network employs view pooling for feature aggregation, which ignores the local view-relevant discriminative information within each view image and the global correlative information across all view images. To leverage both types of information, we propose two self-attention modules, namely, View Attention Module and Instance Attention Module, to learn view and instance attentive features, respectively. The final representation of a 3D object is the aggregation of three features: original, view-attentive, and instance-attentive. Furthermore, we propose employing the ArcFace loss together with the cosine-distance-based triplet-center loss as the metric learning guidance to train our model. As the cosine distance is used to rank the retrieval results, our angular metric learning losses achieve a consistent objective between the training and testing processes, thereby facilitating discriminative feature learning. Extensive experiments and ablation studies are conducted on four publicly available datasets on 3D object retrieval to show the superiority of the proposed method over multiple state-of-the-art methods. 

Multi-view 3D object retrieval leveraging the aggregation of view and instance attentive features

Even though deep face recognition is extensively explored and remarkable advances have been achieved on large-scale in-the-wild dataset, disguised face recognition receives much less attention. Face feature embedding targeting on intra-class compactness and inter-class discrepancy is very challenging as high intra-class diversity and inter-class similarity are very common on the disguised face recognition dataset. In this report, we give the technical details of our submission to the DFW2019 challenge. By using our RetinaFace for face detection and alignment and ArcFace for face feature embedding, we achieve state-of-the-art performance on the DFW2019 challenge.

/pdf/arcface-for-disguised-face-recognition-4dn6o5zwf9.pdf

Stefanos Zafeririou

Papers

ArcFace for Disguised Face Recognition