Kowloon, In Memoriam
Convolutional Neural Networks are tools that can be used to achieve interesting, vague, and impossible graphic scenarios within raster images.
The most well-known application for this type of technology is image classification, creation, and the phenomena known as "deepfakes."
This project aims to utilize the "fuzzy" computational power of neural networks to create a "deepfake" space from existing photographs, images, and drawings.
Kowloon, a now-demolished city in China, was the scene in which these neural networks were deployed.
This is due to the extreme amount of intricacy, resolution, and non-standardization already in place within the city.
Within this project, the neural network generates electrical, plumbing, and mechanical systems.
The goal is not efficiency, rather the creation of ornamental, intricate, and aesthetic systems.
The issue of authorship also impacts the final result of space, as these deepfake images had to be interpreted from a raster format to a vector, 3-dimensional format.
This project was the culmination of a two-semester-long interest in deepfakes, Kowloon, and the possibilities of neural networks.
year: 2020
Credit: Greg Girard and Ian Lambot and the team behind City of Darkness for the
fantastic photographs used in the creation of this project.
