Method and device for multi-channel face liveness detection based on neural network

Abstract

The present invention relates to a neural network-based multi-channel human face living body detection method and device. Its human face living body recognition model not only inputs X human face enlarged image and Y human face detailed image into X+Y channels simultaneously, but also considers two channels through superposition and combination. The fusion of the above-mentioned ones is used to obtain the X+Y+Z human face liveness recognition uncertainty, and determine whether the face image to be detected is a live body. Since the enlarged image of X face and the detailed image of Y face are considered at the same time, as well as the superposition and combination of the internal features of the two, the fused features contain global and local information, which is more conducive to subsequent classification and discrimination; In the training and learning of the recognition model, gradient feedback will promote the classification ability of parallel X+Y channels; compared with the existing technology, the accuracy and accuracy of face recognition are higher, which can further prevent photos, videos, 3D, etc. Various forms of non-biological living body attacks further protect personal and property safety.

Description

technical field [0001] The present invention relates to a deep neural network, in particular to a neural network-based multi-channel live face detection method, device, terminal equipment and computer-readable medium. Background technique [0002] Face, as the most influential biological feature of human beings, has been widely used in all walks of life in authentication systems (as large as access control systems, business systems, payment systems, criminal identification in confidential places, as small as mobile phones, computers and other terminal devices) login to unlock). In particular, with the rapid development of artificial intelligence technology, computer technology, and image recognition technology in recent years, the accuracy of face recognition and face detection has been greatly improved. , it is easy to be attacked by criminals, causing serious personal and property losses. [0003] According to incomplete statistics, the common forms of face non-living at...

AI Technical Summary

Problems solved by technology

[0007] Based on this, it is necessary to provide a multi-channel human face detection method based on neural network for the technical problem that the detection accuracy needs to be improved in the existing human face detection technology, including:

Images