January 11, 2024
VQGAN generated videos

Almost a year ago I’ve been playing around with the VQ-GAN (Vector Quantized Generative Adversarial Networks) model to see if I could generate some interesting, psychedelic visuals. For those unfamiliar with VQ-GAN, it’s a type of generative model that can be used to create high-resolution images. It works by first generating a low-resolution version of the image, then using a discriminator to guide the generation of a higher resolution version.

One of the things that I really like about VQ-GAN is the ability to experiment with different vector quantization settings. By adjusting these settings, I was able to generate a wide variety of different images, ranging from more traditional, photorealistic images to more abstract, surreal compositions.

In the end, I was really happy with the results I was able to achieve using VQ-GAN. It was a lot of fun to experiment with different settings and see what kind of visuals I could come up with, and I’m looking forward to continuing to play around with this model in the future.

Leave a comment

Back To Top