Face deception detection system adversarial sample generation method based on an adversarial generative network

Abstract

The invention discloses a face deception detection system adversarial sample generation method based on an adversarial generative network, and the method can achieve the quick generation of adversarial samples in a large scale according to an original sample image through employing the adversarial generative network technology. Meanwhile, the glasses shape mask is utilized, so that the added countermeasure disturbance is limited in the range of the glasses shape, the actual physical attack can be conveniently carried out in subsequent actual manufacturing, and the countermeasure sample has higher practical value. On the other hand, different face deception detection networks or traditional methods are accessed to the overall training framework, so that adversarial samples for different detection methods can be conveniently generated. The problem that the face spoofing detection system lacks enough adversarial samples in the training process is effectively solved. The confrontation samples of the face spoofing detection method can be automatically generated on a large scale, the difficulty of obtaining a large number of confrontation samples for network training is reduced, and thesafety and reliability of the face spoofing detection method and the face recognition system are improved.

Description

technical field [0001] The invention relates to the technical fields of computer vision and artificial intelligence, in particular to a method for generating an adversarial sample for a face spoofing detection system based on an adversarial generation network. Background technique [0002] Face recognition technology has the characteristics of non-intrusive identity and interactivity. It is more and more widely used in user identity authentication. The technology associated with face recognition has also become a popular research direction in the field of computer vision. But at the same time, with the development of technology, the risk of face recognition system being attacked by spoofing is increasing day by day. In order to ensure the reliability of the face detection system, more and more attention has been paid to face spoofing detection technology. In recent years, with the development of artificial intelligence technology, adversarial attacks against face spoofing d...

AI Technical Summary

Problems solved by technology

However, for face spoofing detection, no large-scale generation of adversarial examples has been reported

This makes it difficult for human deception detection methods to obtain a large number of adversarial samples for training, and cannot effectively resist such adversarial sample attacks.

Images