Face spoofing detection method and system based on multi-scale illumination invariance textural features

Abstract

The invention discloses a face deception detection method and system based on multi-scale illumination invariance textural features. The method comprises the following steps: framing a video, extracting a face image, and carrying out channel separation to obtain a color channel graph; obtaining an illumination invariant texture feature map through an illumination separation texture reservation module, performing normalization, combining the normalized illumination invariant texture feature map with the color channel map to obtain face features, and performing data enhancement to obtain to-be-trained input features; constructing a multi-scale texture module by using the center difference convolution of multiple receptive fields, and embedding the multi-scale texture module into a lightweight network to construct a lightweight multi-scale texture network; weighting the pixel-level loss and the cross entropy loss into total loss; inputting the input features into a lightweight multi-scaletexture network to learn deception features of texture naturalness; updating network parameters according to the loss function, and storing a network model and the parameters after training is completed; and predicting a classification result according to the stored model. The deception features of texture naturalness are accurately extracted, the generalization performance of the model is effectively improved, and the storage and calculation consumption of deployment is reduced.

Description

technical field [0001] The invention relates to the technical field of face spoofing detection, in particular to a face spoofing detection method and system based on multi-scale illumination invariant texture features. Background technique [0002] Face biometrics has rich and unique personal information for authentication and identification, and has become the most popular biometric feature because of its convenience and friendliness, and has achieved extremely accurate identification verification performance. However, there are many kinds of face spoofing attacks: photo attack means that the attacker uses printed photos or face images on the display screen to deceive the authentication system; video replay attack means that the attacker uses the pre-recorded video of the victim to deceive the authentication system ; Face mask attack means that the attacker wears a mask carefully crafted according to the face of the victim to deceive the system; adversarial sample attack me...

AI Technical Summary

Problems solved by technology

These face spoofing attacks are not only low in cost but also can deceive the system, seriously affecting and threatening the application of face recognition systems

[0003] In related research, texture features such as LBP, HoG, SIFT, and SURF are not only relatively rough in texture details but also easily affected by lighting and scenes; the calculation of detection algorithms based on physiological features such as depth maps and rppg signals (facial blood vessel beating signals) is complex non-illumination features such as MSR (multi-scale retinal features) and reflectance maps cannot retain rich deception traces and textures; existing methods generally have large limitations in generalization performance and computational complexity

[0004] The existing face anti-spoofing detection algorithm can achieve a good detection effect in the library, but the generalization performance of the algorithm will be seriously reduced when the illumination and environment change greatly; in addition, the face recognition system is widely used in embedded terminals , mobile devices and monitoring devices, it is difficult to meet the high storage and high computing requirements of existing face anti-spoofing detection algorithms in terms of computing and storage resources

Images