Face exchange method

Abstract

The invention discloses a face exchange method, which comprises the following steps of: inputting a source image and a target image, and acquiring face feature points by using a detector; adjusting the shape of the 3D face through the obtained face feature points; performing face segmentation on the adjusted 3D face shape by adopting an FCN detection model after sample training, and segmenting faces in the source image and the target image from background and occlusion; the face of the target image is effectively covered with the face of the source image, and fusion is carried out; and outputting the picture after face exchange. According to the method, a standard segmentation network obtained by training on rich and diverse samples is used; the FCN realizes accurate and rapid segmentationof the face from the background and the occlusion, and efficiently processes face alignment, face segmentation, 3D shape estimation, expression estimation and the like without any constraint on the image, so that face exchange is more real, and the face change and the recognition rate are reduced.

Description

technical field [0001] The invention relates to the technical field of face detection, in particular to a face exchange method. Background technique [0002] Face swapping was realized as a fully automatic technology more than ten years ago. Its original purpose was to respond to privacy protection issues: it is used to process photos that can be seen everywhere in the public network environment. Face swapping can make the characters appearing in it Features become obscured as a substitute. Since then, however, many face-swapping applications have come more from pastime or entertainment. [0003] Most of the existing face-changing systems have several key aspects in common. [0004] (1) Many methods limit the target photos used for conversion. Given a source image, select a target image that is relatively easier for face exchange from a large face set, such as the target image and the source image. Appearance, including facial tone, posture, expression, etc. Although the...

AI Technical Summary

Problems solved by technology

[0006] (3) Deep learning is often used to deal with face transformation, just like style transformation between images. This method may require training a separate network for each source image, which is impractical for most applications.

[0007] However, when face swapping occurs between randomly selected, unlimited two photos, there is no guarantee of similarity between perspectives, expressions, 3D face shapes, gender, or other attributes that would make the swap easier

Images