LEARNING METHOD AND DEVICE OF GENERATIVE ADVERSARIAL NETWORK FOR CONVERTING BETWEEN HETEROGENEOUS DOMAIN DATA

摘要

According to an aspect of the technical concept of the present disclosure, as a GAN learning method for performing transformation between heterogeneous domains, at least one unpaired training comprising a training image of a first domain and a training image of a second domain performing learning of the first GAN and the second GAN using the data set; converting a first image of the first domain into an image of the second domain using the learned first GAN; re-converting the transformed image of the second domain into a second image of a first domain using the learned second GAN; and performing segmentation of at least one of the first image of the first domain, the image of the transformed second domain, and the second image of the retransformed first domain.