Face anti-counterfeiting detection method, system and device based on cross-modal transformation assistance

Abstract

The invention belongs to the field of image processing and pattern recognition, particularly relates to a human face anti-counterfeiting detection method, system and device based on cross-modal conversion assistance, and aims to solve the problem of low detection performance caused by incapability of utilizing IR modal assistance to perform human face detection under the condition of only configuring an RGB camera. The method comprises the following steps: carrying out face detection and image cutting on an input image; adopting an improved CycleGAN to perform training of a face modal conversion unit through RGB and IR modal training samples, and converting an input image into an IR image; training a face anti-counterfeiting detection model through RGB and IR modal training sample pairs and sample labels, and calculating the probability that an input image belongs to a real face in combination with the input image and an IR image; wherein the face image with the probability greater than the set threshold is a real face image, otherwise the face image is an attack image. According to the method, the anti-counterfeiting information with a more discriminative IR mode is used for assisting the learning of the RGB sample, the IR mode only participates in the training stage, and the model is high in face anti-counterfeiting detection accuracy and low in cost.

Description

technical field [0001] The invention belongs to the field of image processing and pattern recognition, and in particular relates to a face anti-counterfeiting detection method, system and device based on cross-modal transformation assistance. Background technique [0002] The face recognition system has become an indispensable device in daily life, and is widely deployed in mobile phones or scenes equipped with face recognition systems. However, these face recognition systems are extremely vulnerable to malicious attacks by lawbreakers, such as displaying the real user's face image in front of the collection device, attacking samples through video playback or masks, and pretending to be legitimate users to steal private data or money. Therefore, deploying the face anti-counterfeiting detection module on the face recognition system not only has practical application significance, but also has been applied in various scenarios, such as human-computer interaction scenarios such...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the above-mentioned problems in the prior art, that is, the prior art cannot use the IR modality to assist face anti-counterfeiting detection under the condition of only equipped with an RGB camera, thus resulting in the problem of low detection performance. A face anti-counterfeiting detection method assisted by modal transformation, the face anti-counterfeiting detection method comprising:

Images