Face image detection method, device, medium and electronic equipment

Abstract

The embodiment of the invention provides a face image detection method, a device, a medium and electronic equipment; the method comprises the following steps: obtaining an original face image; sampling the original face image according to the preset sampling rules, thus obtaining different sampling signals of the original face image in corresponding different frequency ranges; obtaining the targetface image according to the obtained different sampling signals; using the obtained target face image as the input a pre-trained convolution nerve network classification model, using the convolutionnerve network classification model to detect the target face image, thus obtaining a value outputted by the convolution nerve network classification model and corresponding to the target face image; determining whether the value is smaller than the first preset threshold or not; if yes, determining the original face image as an invalid face image. The method can accurately identify the face imageformed by remaking videos, thus preventing the video remaking attacks, and improving the face image detection accuracy.

Description

technical field [0001] The present invention relates to the technical field of image processing, and in particular to a face image detection method, device, medium and electronic equipment for video reproduction. Background technique [0002] Face recognition technology has been widely used in various fields such as finance, insurance, security, and management to realize the legality verification of customer identities. However, due to the security loopholes in the existing face recognition technology, violations of laws and regulations by counterfeiting customers began to emerge in an endless stream. [0003] At present, the commonly used deception methods for face recognition systems mainly include three categories: 1) photos of legitimate users; 2) videos of legitimate users; 3) 3D models of legitimate users. [0004] Making 3D models of legitimate users is costly, complicated, and difficult to implement, and is rarely used in practical applications. In contrast, photos ...

AI Technical Summary

Problems solved by technology

[0006] However, the inventors of the present invention have found that since the video of the legitimate user may contain the physiological information of the legitimate user's head movement, facial movements, etc., using the physiological information of the legitimate user's head movement, facial movements, etc. It can be verified that the current liveness detection technology is passed, that is to say, the current liveness detection technology cannot recognize the face image obtained based on the legitimate user's video well, and the face recognition system has a security risk of attacking by video remake , the detection accuracy of face images needs to be improved

Images