These three compositions were generated using a GAN trained to come up with believable and stimulating 2D compositions. The composition was represented as a 2D matrix. This was then applied as a mask over a neural synthesis program (credit: Gene Kogan). The generated output was an exploration of how deep learning techniques may be applied to isolated steps in the process of producing 2D compositions.
2017 and Previous
These two pieces were created as a series in one of my final undergraduate exhibits. They're explorations of human "spaghettification" or the process of being sucked into a black hole. This theme was the backdrop for an artistic exploration of isolation and dissociative experiences. These paintings were the results of several months of experimenting with various combinations of compositions and material application. I resurfaced this work when I was producing another generative art project in 2018.