Silent living body detection method, device and equipment and storage medium

Abstract

The invention relates to the field of biological recognition, and discloses a silent living body detection method, device and equipment and a storage medium. The silent living body detection method comprises the following steps: preprocessing a pre-collected training sample image to generate a training sample set; the method comprises the steps that an Xception model is constructed, the Xception model sequentially comprises an input layer, a first convolution layer, a second convolution layer, a global average pooling layer and a binary classifier, and the first convolution layer is composed of 1 * 1 convolution kernels; training the Xception model by adopting the training sample set to obtain a silent living body detection model; collecting a sample image of a to-be-detected object, inputting the sample image into the silent living body detection model for identification, and outputting a silent living body probability corresponding to the to-be-detected image; and if the silent living body probability is greater than a preset threshold, determining that silent living body detection is passed. According to the method, the first convolution layer formed by the 1 * 1 convolution kernel is added on the basis of the Xception model, so that the living body detection accuracy is improved.

Description

technical field [0001] The present invention relates to the field of biological identification, in particular to a silent living body detection method, device, equipment and storage medium. Background technique [0002] With the development of technology, face recognition technology has been widely used in all aspects of people's daily life. In order to avoid offensive login methods of photos or videos, there is a proprietary method in the field of face recognition called "living detection". Common liveness detection methods include cooperative livingness, silent livingness, and texture feature recognition. Among them, the cooperative living body needs to extract specific motion information of the face area for recognition, such as blinking, mouth movement, nodding and shaking the head. However, this requires the cooperation of the user and takes a long time. The texture feature recognition is generally based on the RGB texture features of the image, and combined with the...

AI Technical Summary

Problems solved by technology

However, since the recognition of the face area is interfered by factors such as illumination and occlusion in silent living detection, there is a problem of low recognition accuracy.

Images