Face living body detection method based on transfer learning

Abstract

The invention relates to a face living body detection method based on transfer learning, and belongs to the technical field of image processing and computer vision. The method comprises the followingsteps: segmenting video data into an image sequence, detecting faces in the image sequence, and dividing the data into a training set and a test set; Training the 3D convolutional neural network by using the training set of the source domain to obtain a label classifier for distinguishing true and false faces; Adding a gradient inversion layer behind the convolution layer, and extracting common features of a source domain and a target domain; Performing confrontation training on the data of the source domain and the target domain through a gradient inversion layer to obtain a domain classifierfor distinguishing the data of the source domain and the target domain; And sending the test set of the target domain into the trained label neural network, and selecting the maximum probability of network classification as the final detection result. According to the method, the idea of resistance transfer learning is applied to in-vivo detection, so that the generalization ability of in-vivo detection is improved; Through the 3D convolutional neural network, the spatial information and the time information of the video can be utilized, and the living body detection precision can be improved.

Description

technical field [0001] The invention belongs to the technical field of image processing and computer vision, and relates to a face detection method based on migration learning. Background technique [0002] Currently, biometrics provide a convenient solution for authentication procedures. Face biometric recognition technology, due to its outstanding advantages such as fast detection speed, good user experience, and non-contact, has now been widely used in all aspects of daily life. The existing face biometric technology has great disadvantages and is very easy to be attacked. Common face attack methods include photo attack and video attack. Photo attacks have real face features, while video attacks have more dynamic features of legitimate users, which are more deceptive and seriously affect the accuracy of face recognition systems. [0003] At present, there are three main methods of liveness detection: liveness detection based on texture information, face liveness detect...

AI Technical Summary

Problems solved by technology

The face detection method based on motion information is easily affected by the natural environment such as lighting, and usually needs to process sequence images, consumes a lot of resources, and requires the user to perform specific actions, which is too demanding for the user and the user experience is not good.

[0006] The third is the face detection method based on deep learning: this type of method aims to fully extract the information about the face in the photo, this method has high accuracy on a single data set, but the generalization ability is poor

This is because when using a deep network to learn face features, too much consideration is given to other interference information in the photo, resulting in network overfitting

[0007] The disadvantages of the above three main solutions to liveness detection are that the generalization ability is too poor, and the model trained on one data set cannot be applied to other related data sets.

Images