Head posture deviation correction-based counterfeited face change enhancement detection method

Abstract

The invention discloses a forged face change enhancement detection method based on head posture deviation correction. The method comprises the following steps: acquiring a normal face image and a forged face image and extracting key points; for the forged face image, combining with the head three-dimensional model, carrying out re-projection on the key points to obtain re-projected key points, correcting the projected key points, and according to the extracted initial key points and the corrected key points, carrying out distortion on the forged face image based on a least moving square algorithm to obtain an enhanced forged face image; taking the normal face image and the enhanced counterfeit face image as samples, performing head posture evaluation based on key points of the samples to determine a rotation matrix and a translation matrix of the head posture, and taking the rotation matrix and the translation matrix as construction features to optimize counterfeit face change detection model parameters; and performing enhanced detection of fake face change by using the parameter-optimized fake face change detection model. According to the method, the detection capability of a forged face change detection model is enhanced by enhancing a forged sample.

Description

technical field [0001] The invention belongs to the field of image recognition, and in particular relates to a counterfeit face-changing enhancement detection method based on head posture deviation correction. Background technique [0002] Deepfake videos of human faces are mainly divided into four categories, namely face swapping, face synthesis, facial attribute manipulation and facial expression swapping. Face exchange is mainly to generate a fake A with the same facial expression as A, but replaced by B face; face synthesis is to directly generate a fake face with no clear target through learning; facial attribute operation is similar to beauty The image camera generates some camouflage-like decorations on the human face; the facial expression exchange is to apply B's expression to A, so that the original facial expression in A is consistent with B. [0003] The face exchange technology mainly relies on the deep learning network and the confrontation generation network ...

AI Technical Summary

Problems solved by technology

[0007] Although the above two existing documents use deep learning methods to detect fake face images, they are not suitable for the detection of all fake faces because only the visual biometric features are considered.

Images