Use AI to write the Wesleyan score

For his senior citizens' work, Ellington Davis '25 set out to create a musical score of Wesleyan, which consisted of the sounds that make the campus known. He defined and isolated the chirping birds, heating units, conversations between classmates and remote sirens from passing emergency vehicles. Then he took his project one step further with contemporary tools such as artificial intelligence.

Davis, studying music and science and technology, formed a mechanical algorithm with open source software to meet the application for sounds and images that were collected for his senior thesis project “Wipeople: The Wild, Wild Urban”. For example, if an subject moved an arm in front of the movement camera used, the AI ​​model found a sound and a picture that was closest with the movement. Then the system projected the corresponding image on the screen and played the corresponding sounds loudly.

To train the model, he initially spent hours going through the campus and listening to latent noises. After he had the feeling that he had heard enough, he began to record and photograph what he was watching. He documented a wide range of inputs on campus and outside, from the construction of the new science building to steam systems in the Freeman Athletic Center to a biofuel plant in Southington, Conn. This processes Wesleyan's food waste.

“I think what was really exciting to walk around and document it is that it is a way of seeing her space,” said Davis. “If you can listen and hear things and ask yourself, why do these things happen and where do you come from? It is also, how does the thing actually come? Sound is a way to question and explore.”

Next, he turned an integer to each input to serve as an anchoring point for the software in its score. “[The program] Can transform movement into data and make sense very easily, “said Davis.

After completing the completion, Davis demonstrated his creation in a concert with his band STATIC on April 12th in the Hewitt workshop.

“This system was able to create a living, constantly changing, not sovereign soundscape, based on the score of movement in the room that made it possible through training,” wrote Davis in his thesis paper. “The webcam (from my computer comes from my computer) read the audience that took part in my diploma thesis and Hewitt workshop as a score – and overrun every participant into a sound walker.”

His project was inspired by ReadyMade art practices in experimental music, such as the late John Cage, an influential experimental composer who was involved in the music department of the university for over 50 years, and his score “Atlas Eclipticalis”, which overlay musical staves via Star chart. Davis is also interested in the “tendency of urban people, people who can be considered everyday as everyday, he said. He wanted to understand the entries that we may take granted. For this purpose, he used anthropological and musical drawing systems in the form of his spacers and semiotics.

“I wanted to create a system that could compress all of this infrastructure that is part of this disappointed landscape, and make it a kind of enchantment from it in which they are not able to ignore all of these noise at once,” said Davis.

After graduating, Davis plans to return to Bay Area to work with political organizations and work in urban farms.