"The biggest artistic achievement of the year" - Artsy

CAN: Creative Adversarial Networks: Generating "art" by

 Learning about Art and Deviating from Style Norms

If we teach the machine about art and art styles and force it to generate novel images that do not follow established styles, what would it generate? Would it generate something that is aesthetically appealing to humans? Would that be considered “art”? read more here

Our study’s emphasis is on understanding how the machine achieves classification of style, what internal representation it uses to achieve this task, and how that representation is related to art history methodologies for identifying styles.  read more here

"The art challenges technology and the technology inspires the art"
                                                                                                    -- John Lasseter

Can we measure creativity of paintings, only based on their visual characteristics? 

Having a large collection of fine-art paintings, we would like to know which paintings were more creative in the eyes of the machine. We propose a computational framework that can compare paintings based on their creativity over time. Here you can find more detail about this project.  

Can find an influence path in a network of artist, when we only have images of paintings ? 

In this project, we proposed an algorithm to find influence paths based on visual similarity of paintings (without human supervision) and verified our results with what Art historian find out. 

How can we measure similarity of two paintings visually? Which visual features are better that others? How can we learn the right data-driven metric with minimum human supervision?

Knowledge discovery in collection of paintings would not be possible without having a metric to measure visual similarity between pairs of items. In this project, we proposed a metric that is uniquely designed for fine-art paintings.