Face anti-spoofing based on color texture analysis

By Zinelabinde Boulkenafet, Jukka Komulainen, Abdenour Hadid. International Conference on Information Photonics, 2015

433 citations recorded by the source corpus.

Research on face spoofing detection has mainly been focused on analyzing the luminance of the face images, hence discarding the chrominance information which can be useful for discriminating fake faces from genuine ones. In this work, we propose a new face anti-spoofing method based on color texture analysis. We analyze the joint color-texture information from the luminance and the chrominance channels using a color local binary pattern descriptor. More specifically, the feature histograms are extracted from each image band separately. Extensive experiments on two benchmark datasets, namely CASIA face anti-spoofing and Replay-Attack databases, showed excellent results compared to the state-of-the-art. Most importantly, our inter-database evaluation depicts that the proposed approach showed very promising generalization capabilities.

Subfields: Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/9c9edd91d021ed05ef4a16890df83a25064b1c77

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

Thin-Plate Spline Motion Model for Image Animation (2022, methodology): “While the proposed method may be used to make fake videos for spoofing, some detection software will easily determine the authenticity of videos by analyzing the color textures [3] or using the depth information obtained through the proximity sensors, which cannot be simulated by 2D image generating”
Face Spoofing Detection Using Colour Texture Analysis (2016, background): “This present work extends our preliminary colour texture based approach presented in [15] and provides an in-depth analysis on the use of colour texture analysis for face spoofing detection.”
Searching Central Difference Convolutional Networks for Face Anti-Spoofing (2020, methodology): “Several classical lo-cal descriptors such as LBP [7, 15], SIFT [44], SURF [9], HOG [29] and DoG [45] are utilized to extract frame level features while video level methods usually capture dynamic clues like dynamic texture [28], micro-motion [53] and eye blinking [41].”
Face anti-spoofing using patch and depth-based CNNs (2017, methodology): “On the other hand, HSV and Y CbCr are based on the separation of the luminance and the chrominance information, providing additional features for learning the discriminative cues.”
Face Anti-Spoofing with Human Material Perception (2020, methodology): “Several classical lo-cal descriptors such as LBP [2,4], SIFT [7], SURF [19], HOG [5] and DoG [6] are utilized to extract frame level features while video level methods usually capture dynamic cues like dynamic texture [20], micro–motion [21] and eye blinking [22].”
Face De-Spoofing: Anti-Spoofing via Noise Modeling (2018, background): “To overcome some of these difficulties, researchers tackle the problem in different domains, such as HSV and YCbCr color space [24,25], temporal domain [26,27,28,29] and Fourier spectrum [30].”
Deep Spatial Gradient and Temporal Depth Learning for Face Anti-Spoofing (2020, background): “where K i is the ith contrastive convolution kernel, i ∈ [0, 7].”
Deep Learning in the Wild (2018, neutral): “These techniques mainly focus on texture differences between real and spoofed images, differences in color space [7], Fourier spectra [30], or optical flow maps [6].”
Secure Face Unlock: Spoof Detection on Smartphones (2016, neutral): “%) and [14] (2.9”
Deep Tree Learning for Zero-Shot Face Anti-Spoofing (2019, methodology): “Limitation of feature selection: They use handcrafted features such as LBP to represent live faces, which were shown to be less effective for known spoof detection [27, 32, 37, 48].”
Face Anti-Spoofing: Model Matters, so Does Data (2019, background): “investigated different color spaces such as HSV and YCbCr [6, 8].”
Learning Deep Models for Face Anti-Spoofing: Binary or Auxiliary Supervision (2018, methodology): “Many handcrafted feature-based methods may improve intra-database testing performance by simply concatenating the features of consecutive frames to train the classifiers [11,18,28].”
In Ictu Oculi: Exposing AI Generated Fake Face Videos by Detecting Eye Blinking (2018, methodology): “Detecting eye blinking has been studied in computer visions for applications in fatigue detection [10]–[14] and face spoof detection [15]–[19], and various methods have been proposed”
Cross-Domain Face Presentation Attack Detection via Multi-Domain Disentangled Representation Learning (2020, methodology): “…that leads to poor generalization ability of PAD models under new application scenarios. face PAs, which assume that there are inherent disparities between live and spoof faces, such as texture in color space [4, 6], image distortion [48], temporal variation [39] or deep semantic features [50, 36].”
In Ictu Oculi: Exposing AI Created Fake Videos by Detecting Eye Blinking (2018, background): “Detecting eye blinking has previously been studied in computer visions for applications in fatigue detection [14, 34, 8, 2, 22] and face spoof detection [4, 10, 20, 30, 19].”
OULU-NPU: A Mobile Face Presentation Attack Database with Real-World Variations (2017, methodology): “As a baseline face PAD method, we chose the color texture based method [2] as it has shown promising generalization abilities.”
Animate Anyone: Consistent and Controllable Image-to-Video Synthesis for Character Animation (2023, methodology): “The proposed method may be used to produce fake videos of individuals, which can be detected using some face anti-spoofing techniques[5, 51, 58].”