The process involves the use of a variant of Generative Adversarial Networks or GANs called the Gated-GAN which involves adversarial training of the model to generate convincing counterfeits. The idea is extract the features and content of an input image and render it as different artistic styles (e.g. famous paintings , artworks etc.)
Generative Network
Derived from : Gated-GAN: Adversarial Gated Networks for Multi-Collection Style Transfer Xinyuan Chen, Chang Xu, Xiaokang Yang, Senior Member, IEEE, Li Song, and Dacheng Tao)
The generative networks have three modules: an encoder, a gated transformer, and a decoder. Different styles can be achieved by passing input images through different branches of the gated transformer.
Base Images
Style Images
Output Images
Discriminative Network
Derived from : Gated-GAN: Adversarial Gated Networks for Multi-Collection Style Transfer Xinyuan Chen, Chang Xu, Xiaokang Yang, Senior Member, IEEE, Li Song, and Dacheng Tao)
The discriminative networks are used to distinguish whether the input image is a stylized or genuine image. An auxiliary classifier is used to recognize the style categories of transferred images, thereby helping the generative networks generate images in multiple styles.
Hosier Lane, Melbourne
Sydney Harbour Bridge, Sydney
Opera House, Sydney
Yarra River, Melbourne