Method, system and equipment for face-changing forgery detection based on three-dimensional shape of human face

Abstract

The present invention belongs to the fields of face swapping images, computer vision and image forensics, and specifically relates to a method, system and equipment for detecting face swapping forgery based on the three-dimensional shape of the face, aiming to solve the problem that the prior art is susceptible to post-processing operations and non-identical For the attack of the source face forgery method, the robustness and generalization of the model are not high, so the detection accuracy of face forgery is low or even impossible to detect. The invention includes: extracting the facial appearance of the person to determine the identity of the person; obtaining the three-dimensional facial shape feature template of the corresponding person from the registration reference set; extracting the three-dimensional facial shape feature of the image to be detected through 3DMM; calculating the Mahalanobis distance between the two, if less than True or false judgment threshold, it is a real face image; otherwise, it is a forged image with face swapping. The invention can detect forged samples from real human face pictures, and effectively improve the anti-interference ability of the detection model for various post-processing operations and the generalization ability of forgery methods that are not of the same source as training data.

Description

technical field [0001] The invention belongs to the fields of face-swapping images, computer vision and image forensics, and in particular relates to a face-swapping forgery detection method, system and equipment based on the three-dimensional shape of a human face. Background technique [0002] In recent years, with the vigorous development of deep learning, especially the generative adversarial network (GAN), the face tampering technology of face images has also shown rapid progress. With the development of GAN in recent years, some high-quality fake images that are difficult for human eyes to distinguish can be easily generated. Among them, face swapping is currently the most common forgery generation method. Some criminals can use the forgery generation technology of face exchange to achieve some malicious purposes, so this is also one of the important research topics in the field of digital image forensics. [0003] In the field of digital image forensics, some deep f...

AI Technical Summary

Problems solved by technology

[0011] In order to solve the above-mentioned problems in the prior art, that is, the prior art is vulnerable to post-processing operations and non-homologous face forgery methods, and the model robustness and generalization are not high, so that the accuracy of face forgery detection is relatively low. low or even impossible to detect, the present invention provides a face-changing forgery detection method based on the three-dimensional shape of the face, the method comprising:

Images