We are proud to announce this big art exhibition opening and demos of some of the projects collaboratively created in our Artificial Intelligence and Art Pavilion at CCH.
This is an art project created by internationally famous Spanish artist Sergio Albiac. It is an interactive live portrait art installation which uses generative code and deep neural networks to move us into a symbiotic relationship between others and the environment. The artist questions the use of algorithms not only as a source of optimization of time, money, or resources, but as a critical thinking tool.
A music video was created with artificial intelligence. A deep structure (StyleGAN2) was first learned based on thousands of images collected online which represent the conceptual meaning of the song’s lyrics. The resulting artificially generated video guides the viewer through the multi-dimensional latent space of neural structures in the form of a journey through a changing landscape of newly created artworks.
The smart photobooth
The idea of the PhotoBooth is to transform a photo taken by the visitor into a preselected artistic painting style. The visitor is guided in natural language by Cutie, our QT robot.
AI Avatar Dancer
In this project, our objective was to generate a classical dance sequence based on the “Dying Swan” music piece, using the Dancing2Music dance generator. This work has been conducted in AI Robo-lab at the University of Luxembourg. In collaboration with Betania Antico ( dancer and choreographer ), we also applied the Open-Pose pose detection model on a music video of Natalia Petrovna Osipova (principal ballerina with the Royal Ballet in London) performing the “Dying Swan” to analyze and compare the AI-generated dance movement and human dance movements from a choreographic
point of view.
“Interactive Mirror” project
The aim of the art installation “Mirror, Mirror” is to transform a person’s image in order to prompt them to examine their version of self and the surrounding reality and make them aware of the developments in the field of Artificial Intelligence. To provide a unique experience of interacting with the paintings, in our installation, we use deep fake, neural style transfer and living portraits techniques.
Duchampian art made by AI DALL-Emades experiments with cutting-edge ML models like DALL-E, ruDALL-E and CLIP by Open AI to generate images of mysterious objects resembling 20th century industrial design with relevant art titles. In this project we are attempting to create a DALL-E like tool or some combination of ‘VAQ + Transformers’ based tool to be utilised by the art community. Considering Marcel Duchamp’s original approach, a machine learning based workflow which combines ‘text-prompt engineering’ plus ‘model-stacking’ method can yield interesting results.
AI Critique project
The AI Critique Project is an attempt to write art criticism using Artificial Intelligence. Our system aims to automatically generate a text to give a review on a work of art, by receiving an image as input and producing a text, in English, as output, which gives a description and interpretation on the work in question. Our solution leverages the recent vision-language pre-training (VLP) approaches, including Object-Semantics Aligned Pre-training (OSCAR) and Visual-Vocabulary pre-training (VIVO), by combining them in a single intelligent art critique that provides accurate descriptions.
Automatic mixed painting
Style Transfer is the technique of applying the style of one image to the content of the other using a deep neural network algorithm. This technique is being extended in this project. The main goal is to impose a painting or a section of it into another painting in such a way that it looks to be genuine. The content object is placed on the style painting, and both content and style are then harmonized by the neural network. It transfers the relevant characteristics of the object onto the target painting and improves the output quality. This project could help in the creation of better artwork with surprising results.
This project deals with the complex topic of AIs as possible superior and almighty entities. Throughout history this role was mostly associated with gods and goddesses. We designed a chatbot which can adopt different identities based on different religious and philosophical scriptures. Users can talk to DEUS directly or witness a conversation between two “divine” chatbot-identities. Following the basic idea of “god as a laughing woman” our approach as a large and intercultural project-team with diverse backgrounds includes philosophical, religious humanistic and also humourful aspects. We want to engage our audiences live and online with many open questions about ethics, religion, self-perception and possible prejudice and bias.
How to access