Face aging processing method and system based on multiple emotions

Abstract

The invention discloses a face aging processing method based on multiple emotions. The method comprises the following steps: acquiring an original picture and a target picture corresponding to the original picture; respectively inputting the obtained original picture and the target picture into encoders Ee and EI of the trained first network model so as to respectively obtain key features of the original picture and key features of the target picture; inputting the obtained key features of the original picture and the key features of the target picture into a generator GE of a trained first generative network to obtain a color picture; and inputting the color picture into a generator GA of a trained second network model to obtain an aged color picture. According to the method and system, multiple losses are utilized to ensure that the CD-Net learns a real image under the condition that other information is not lost, vivid facial aging images under different emotions can be generated, and the emotional characteristics are specific and not random.

Description

Technical field[0001]The present invention belongs to the field of image processing, and more particularly to a multi-emotion-based human face aging treatment method and system.Background technique[0002]The existing human face aging treatment technology mainly includes two processes, which build large-scale human face aging data sets, and building a novel depth learning framework. It is necessary to build a large-scale face aging data set needs to collect face images of each age group, and the new depth learning framework needs to be constructed by building a network model, and the network model is deeply trained.[0003]However, existing human face aging treatment technology still has some irreparable flaws:[0004]First, the current human face aging algorithm is based on a single person's face emotions, and actually different emotional results are different, which leads to poor performance in the current human face aging algorithm, and is difficult to further further Increase;[0005]Se...

AI Technical Summary

Problems solved by technology

[0007] In view of the above defects or improvement needs of the prior art, the present invention provides a face aging processing method and system based on multiple emotions, and its purpose is to solve the technical problem of poor performance of the face aging algorithm in the existing face aging processing method , labeling large-scale data sets requires huge manpower and material support, and the technical problems of collecting images of the same face image in different periods take a long time, and it is difficult to perform aging processing on the specified aging interval

Images