Unsupervised Image-to-Image Translation Based on Bidirectional Style Transfer

Image-to-image translation (I2I) is an image synthesis technique to map a source image to the style of the target domain while preserving its content information. Existing image-to-image translation study results showed excellent image synthesis performance using generative adversarial network (GAN) based models, but they are not capable of efficiently handling the style of the target domain. To overcome this limitation, a bidirectional style transfer network has been developed to perform image synthesis by intersecting images of two domains with each other's styles, but the type of applicable dataset is limited due to supervised learning-based training. We proposed an unsupervised image-to-image translation method by employing a bidirectional style transfer network with a cyclic collaborative loss to train the model. Experimental results showed that the proposed network accurately reflected the style of the target domain in the image synthesis task.