Face image migration method and device

Abstract

The invention relates to the technical field of image processing, and provides a face image migration method and device, and the method comprises the steps: obtaining a face image sample set, a target domain image generator, and a first discriminator; augmenting the clear face image sample to obtain a target domain positive example sample; predicting a first clear face image of the shielded face image sample in a target domain through a target domain image generator, and taking the first clear face image as a target domain negative example sample; and taking the target domain positive example sample and the target domain negative example sample as the input of a first discriminator, combining a preset first contrast loss function, performing contrast learning training on the first discriminator, and iteratively performing adversarial training on the target domain image generator and the first discriminator after contrast learning training to obtain a trained target domain image generator. The overall stability of the generative adversarial network is improved, and the performance of the image generator is improved.

Description

technical field [0001] The present application relates to the field of face recognition, in particular, to a face image transfer method and device. Background technique [0002] In the field of face recognition, images where the face is blocked by objects such as masks, mosaic face headshots, and images with damaged and incomplete headshots belong to large-area face occlusion images, and key points such as face position and positioning of facial features are blocked, which affects recognition. Effect. The relevant face occlusion recognition technology requires a lot of cost to improve the information capture ability or requires a large scale of face data sets, and the cumbersome preprocessing work such as manual labeling greatly increases the training cost. Contents of the invention [0003] The purpose of this application is to provide a method and device for face image migration. By performing comparative learning training on the first discriminator, and performing conf...

AI Technical Summary

Problems solved by technology

[0002] In the field of face recognition, images where the face is blocked by objects such as masks, mosaic face headshots, and images with damaged and incomplete headshots belong to large-area face occlusion images, and key points such as face position and positioning of facial features are blocked, which affects recognition. Effect

Relevant face occlusion recognition technology requires a lot of cost to improve the ability of information capture or requires a large scale of face data sets, and the cumbersome preprocessing work such as manual labeling greatly increases the training cost

Images