Towards Face Presentation Attack Detection Based on Residual Color Texture Representation

Abstract

Most existing face authentication systems have limitations when facing the challenge raised by presentation attacks, which probably leads to some dangerous activities when using facial unlocking for smart device, facial access to control system, and face scan payment. Accordingly, as a security guarantee to prevent the face authentication from being attacked, the study of face presentation attack detection is developed in this community. In this work, a face presentation attack detector is designed based on residual color texture representation (RCTR). Existing methods lack of effective data preprocessing, and we propose to adopt DW-filter for obtaining residual image, which can effectively improve the detection efficiency. Subsequently, powerful CM texture descriptor is introduced, which performs better than widely used descriptors such as LBP or LPQ. Additionally, representative texture features are extracted from not only RGB space but also more discriminative color spaces such as HSV, YCbCr, and CIE 1976 L∗a∗b (LAB). Meanwhile, the RCTR is fed into the well-designed classifier. Specifically, we compare and analyze the performance of advanced classifiers, among which an ensemble classifier based on a probabilistic voting decision is our optimal choice. Extensive experimental results empirically verify the proposed face presentation attack detector’s superior performance both in the cases of intradataset and interdataset (mismatched training-testing samples) evaluation.

Myope Models -- Are face presentation attack detection models short-sighted?

“…CoALBP (YCBCR) [20] 17.1 CoALBP (HSV) [20] 16.4 Color [2,7] 13.9 De Spoofing [15,7] 12.3 RCTR-all spaces [7] 10.7 ResNet-18 [9] 9.3 SE-ResNet18 [10] 8.6 AlexNet [20] 8.0 DR-UDA (SE-ResNet18) [29] 8.0 DR-UDA (ResNet-18) [29] 7.2 3D-CNN [19] 7.0 Blink-CNN [8] 4.6 DRL-FAS [3] 1.8…”

“…The variability of attacks included in the dataset significantly increase the difficulty of finding a model capable of performing well on all of them. For this reason, even methods that achieve almost zero error on other datasets, have worse performance on the ROSE-Youtu [7].…”

“…Examples of these images are seen in Figure 4. We used the videos from the first 10 indexed subjects (2,3,4,5,6,7,9,10,11,12) for training and the remaining 10 for testing.…”

Active Face Spoof Detection Using Image Distortion Analysis

“…Deep learning can be used to generate promising results in the field of computer vision, and it has been shown to be useful in solving the challenge of face spoof detection [39]. The CNN architecture employed in this research is as follows: To extract various features from input images, a 1-dimensional convolutional layer (Conv1D) with 32 units as the input layer and an input dimension of (2280 x1) is used.…”

An improved simple feature set for face presentation attack detection

“…The motion based PAD methods suffer from long time of data registration and require in most cases additional actions from the user. The second approach is texture based methods [Pen18;Pen20;Du21]. Texture based methods suppose that the real face image has significantly different textural properties from the fake image reproduced using printing or screen devices.…”