This year the contest RobotArt had 19 teams that delivered more than 100 pieces of art made exclusively by robots, that is to say, without human intervention. Let’s say it is something like a proof of Turing artistic. Can a robot make art? How can he understand the aesthetic values of the works? What does it mean to finally be an artist and devote himself to creating art? For Andy Warhol, “you know something of that art when you go to the bank and you can cash the check for your work”. Let’s look at but the answers are robotic.
First place: CloudPainter ($40,000) – Independent/Gonzaga(USA)
Incorporating technology of machine learning, CloudPainter was able to paint pictures that are evocative, with varying degrees of abstraction. The source code and the 3D models of the program are available online. In addition your creator has written on his philosophy and methods.
This work was created by a robot using multiple neural networks, a variety of the IA, as well as loops of feedback at all levels
Second place: PIX18 / Creative Machines Lab ($25,000) – Columbia University (USA)
A team from the University of Columbia, continuing her work of 2017, follows the line of impressionism, showing all kinds of skills with the brushes, and the strokes.
Nature dead (original)
Muhammad Ali (original Image)
Third place: CMIT ReART ($10,000) – Kasetsart University (Thailand)
(Category: art re-interpreted)
It is clear that the use of new technologies, such as neural networks deep, have opened an interesting field in regard to the manner of interpreting the environment information. It is fair to say that all of these networks, machine learning, draw conclusions from a massive amount of data and it is in some ways a model of how neurons work in the human brain.
There are more prizes, which can be seen here. The question that remains in the air is: what are the robots are making art or just digital processes of images more or less developed?What thinks the reader of unocero?