Descripción

Denbora Latente is a generative art intervention, based on artificial intelligence, which reinterprets the passage of time projected on the façade of the Kursaal. Seconds, minutes, years, in short, time is represented by artificially generated images. In this way, the iconic façade will show a dimension of that time, a reflection of that latent and infinite space provided by the Generative Adversarial Neural Networks (GAN). To create Denbora Latente, an artificial intelligence based on generative adversarial networks has been used, which are composed of two deep neural networks: one that produces images and one that judges them. It is a game in which both networks converse-discuss to create images that they have previously learned to differentiate. One paints images and the other criticises whether the image is appropriate. These GANs use the conversation to generate images and, when the conversation ends, what you get is the ability to produce infinite new images. For example, if we have given the deep network that criticises images of forests to the deep network that criticises, the generating network will offer infinite variations of forests. The façade of the Kursaal will thus be a latent space in which to move between the representations generated in it, with the aim of finding visual patterns that reflect the passage of time.

Temática
Tipo de Obra
Año
2022
Material
Programming code and algorithms
Imagenes
Imagen
Imagen Listado
Imagen
Tipo de proyecto