Estudio de arte digital mediante la aplicación de técnicas de deep learning.

Abstract

This research project will provide information about digital art and pictorial art represented in the area of artificial intelligence through the application of Deep learning techniques to obtain images generated by a neural network through the input of a text as a parameter. This work is focused on a current problem that is related to the rapid and constant growth of artificial intelligence models capable of generating artistic content competing in the near future with the art generated by a human being. Within this pragmatic study we will take artistic references that influence diverse artistic interventions at a global, national and local level. So, are artificial intelligences capable of creating any image that they are asked to write? What would be the consequences for illustrators, graphic designers, painters, photographers, for the standard human being? Will he be able to bring to reality any idea he has in his head and the fear that these AIs will be used for something negative? If this that AIs create can be called art, what will it mean when many humans begin to prefer, experiment, paintings, songs, books created by AIs instead of those created by other humans? For this reason, the artistic object of my present project is to provide knowledge about the arrival of new technological resources such as artificial intelligence and deep learning which help us as a source of support for the creation of new ideas in art, so that new techniques and increasingly accurate models continue to be developed. With which it has been possible to generate images from scratch from the learning of neural networks, adding also the ability to interpret human language from natural language processing techniques.