Anime Style Space Exploration Using Metric Learning and Generative Adversarial Networks
This work addresses the challenge of understanding and manipulating artistic style in image generation, specifically for anime illustrations, though it is incremental in its approach.
The paper tackled the problem of defining and encoding artistic style by proposing a metric learning method to capture style differences between artists, which was validated through classification performance and used to generate style-conditioned anime portraits with a GAN.
Deep learning-based style transfer between images has recently become a popular area of research. A common way of encoding "style" is through a feature representation based on the Gram matrix of features extracted by some pre-trained neural network or some other form of feature statistics. Such a definition is based on an arbitrary human decision and may not best capture what a style really is. In trying to gain a better understanding of "style", we propose a metric learning-based method to explicitly encode the style of an artwork. In particular, our definition of style captures the differences between artists, as shown by classification performances, and such that the style representation can be interpreted, manipulated and visualized through style-conditioned image generation through a Generative Adversarial Network. We employ this method to explore the style space of anime portrait illustrations.