Neuroscience and AI Collide in “Mind to Image” Art Project

The hyper-color image of a black hill and lava spillage is pretty enough, but the fact that AI has digitally created it makes it unique. It results from the brain waves of one French art collective member, Obvious, who contributed his MRI data from the Brain Institute (ICM) at the Pitie Salpetriere Hospital in Paris, France.

AI art from brain waves

The famous three guys, Fautrel, Hugo Caselles-Dupre, and Gauthier Vernier, are aged in appearance. They are making waves across the art world after they sold an Artificial intelligence-generated artwork at Christie’s in New York for more than €400,000.

The “Mind to Image” project used a publicly available MindEye program that can download the nerve activity of the viewed images and re-construct them. Then, their own AI program was incorporated to create artwork.

The team tried two different types of experiments: the first involved visualizing pictures and the brainwaves recorded in the MRI, and the second involved manual reproduction of the observed shapes rather than mental visualization of the shapes.

Neuroscience meets AI creativity

In their problem, they have to apply the machine several times, which needs 10 hours for each data set to teach the AI machine the answers for the given purposes. “Before this, we’ve been running this research for almost 10 years, and we know that it’s possible to create just any visual in someone’s mind through this technique,” said an Abuja-based researcher doctor at the Brain Institute Consortium, Dr. Alizee Lopez-Persemb.

The essence of painting is transcribing a manifest, physical reality into the two dimensions of the canvas. The team might have outwitted if it took hours to complete a neural network-fed MRI study before they could push their own AI program into it and create a vibe laced with a part of Surrealism.

He mentions the progress in the quality of medical imaging, which is way beyond what can be produced in the film, and the breakthrough of generative AI technology, which can generate images from the user’s description.

Caselles-Dupre claimed that very real connections exist between art and science. She highlighted this technology as “scary as it may sound” if used wrongly. The project results will eventually be presented and available to the general public at the Danysz gallery in October. The group is interested in expanding the project into the genre of sound and video.