NEW LENSES: Experimental Capture and Release

This is an extension of the first speed project, where I used AI to generate photographs. This time, I tried to use GauGAN2 to create landscapes using only the text input and the provided filter images, which resulted in pictures of landscapes that don’t exist anywhere.

Originally, this project was going to be about photographs in which objects are replaced with 3D models of themselves, with an appended commentary on how a photograph of an object and a low-poly model of an object are both virtual (unreal) representations of the real thing and aren’t that different despite appearances. However, my attempt to make a 3D model of my coffee failed the first couple of tries, and after waiting an hour when I thought I succeeded I got this:

So it really didn’t work at all, and I decided to just throw out the idea and try something else.