crea.visions Venice 2022

CREA Visions Venice is a collaborative event series between the Center of Hybrid Intelligence (Aarhus, Denmark), and the ACM Creativity & Cognition 2022 conference. Our aim is that the AI-assisted activity that crea.visions allows for can help anyone visualize the future of Venice. We also wish to spark dialogues and foster community action around sustainable development and the future of the city of Venice. 

crea.visions  uses what is called a Generative Adversarial Network (GAN), which is a type of machine learning framework. Since we wanted to create visions of Venice that resemble the city, we built a special version of the algorithm, which was trained on both images of building-like structures from Artbreeder’s website (the project that inspired the creation of crea.visions) and images taken in Venice by our local collaborators.

Further information

Janet Frances Rafner

AIAS-SHAPE Fellow Aarhus Institute of Advanced Studies

In May ’22, we held pop-up events at various locations in the city of Venice. We introduced ourselves and our project and asked people to play with our game for a bit. We talked with local clubs, students, and tourists. Over the course of one week, 4668 images were generated, 235 submitted, and 70 ranked. The top 10 images will be exhibited in Giudecca, owned by Insituzione Fondazione Bevilacqua La Masa in Venice, from 22-25 June, as part of the ACM Creativity & Cognition 2022 Conference Exhibition.

Some pictures taken in Venice.

The images of Venetian buildings were then “translated” into a vector space that the GAN understands, which resulted in images like these.

The winning images