Title: CAN: Creative Adversarial Networks Generating "Art"
Time:5th 11 3:30
Elgammal et al. proposed a new model for generating art. The model uses the Generative Adversarial Networks (GAN) architecture, combined with a new objective function to generate creative art. They arouse the model to generate creative art by maximizing the deviation from existing styles, while minimizing the deviation from art distribution. They used human subjects to show that CAN's generated art is indistinguishable from art created by contemporary artists.
Nadav Schor is currently pursuing his MSc in the fields of Computer Graphics and Machine Learning under the supervision of Prof. Daniel Cohen-Or at Tel-Aviv University.