Abstract
Style transfer describes the rendering of an image semantic content asdifferent artistic styles. Recently, generative adversarial networks (GANs)have emerged as an effective approach in style transfer by adversariallytraining the generator to synthesize convincing counterfeits. However,traditional GAN suffers from the mode collapse issue, resulting in unstabletraining and making style transfer quality difficult to guarantee. In addition,the GAN generator is only compatible with one style, so a series of GANs mustbe trained to provide users with choices to transfer more than one kind ofstyle. In this paper, we focus on tackling these challenges and limitations toimprove style transfer. We propose adversarial gated networks (Gated GAN) totransfer multiple styles in a single model. The generative networks have threemodules: an encoder, a gated transformer, and a decoder. Different styles canbe achieved by passing input images through different branches of the gatedtransformer. To stabilize training, the encoder and decoder are combined as anautoencoder to reconstruct the input images. The discriminative networks areused to distinguish whether the input image is a stylized or genuine image. Anauxiliary classifier is used to recognize the style categories of transferredimages, thereby helping the generative networks generate images in multiplestyles. In addition, Gated GAN makes it possible to explore a new style byinvestigating styles learned from artists or genres. Our extensive experimentsdemonstrate the stability and effectiveness of the proposed model formultistyle transfer.