Training image-to-image translation neural networks

Abstract

A computer-implemented method for training a forward generator neural network G to translate a source image in a source domain X to a corresponding target image in a target domain Y is described. The method includes: obtaining a source training dataset sampled from the source domain X according to a source domain distribution, the source training dataset comprising a plurality of source training images; obtaining a target training dataset sampled from the target domain Y according to a target domain distribution, the target training dataset comprising a plurality of target training images; for each of the source training images in the source training dataset, translating, using the forward generator neural network G, each source training image to a respective translated target image in the target domain Y according to current values of forward generator parameters of the forward generator neural network G; for each of the target training images in the target training dataset, translating, using a backward generator neural network F, each target training image to a respective translated source image in the source domain X according to current values of backward generator parameters of the backward generator neural network F; and training the forward generator neural network G jointly with the backward generator neural network F by adjusting the current values of the forward generator parameters and the backward generator parameters to optimize an objective function, wherein the objective function comprises a harmonic loss component that ensures (i) similarity-consistency between patches in each source training image and patches in its corresponding translated target image, and (ii) similarity-consistency between patches in each target training image and patches in its corresponding translated source image.