A StyleGAN latent space navigator using audio and 2d vector drawing. Builds latent space vectors from multiple audio files concurrently, with a number of options for shaping the audio into the vector space: amplitude, orientation, offsetting, etc.
Uses p5.js and Webpack, requests StyleGAN images from RunwayML models.
npm install
npm run serve
Set w.testRun = false
to effectively pull images from the running RunwayML model.
sample-output-@morekid-fires-test.mov
"Fires" by @morekid.