StyleGAN 2 Truncation
ฝัง
- เผยแพร่เมื่อ 22 ก.ย. 2024
- StyleGAN 2 trained on images of landscapes, with varying levels of truncation.
At the beginning, all images have been fully truncated, showing the "average" landscape of all generated landscapes. When ψ = 1, no truncation is applied, showing full diversity across the distribution of images. When ψ = 1.5, generated images can reach beyond its own distribution, lowering image quality significantly, but increasing diversity beyond what is even in the real distribution.
Code: github.com/man...
Research Paper: Analyzing and Improving the Image Quality of StyleGAN
StyleGAN, Tensorflow 2.0, Truncation, Landscapes
Looking foward for more tutorials of how to use this software
I really want to use it
It was really cool, please explain how to use a pre-trained network?
Hello. Thank you very much for your videos and implementations of StyleGAN 2 Tensorflow 2 on github. Could you tell us more about what computer equipment you use for training? How long did it take you? How do you collect data for training?
Hey, thanks for the support! I'm using a single computer with a GTX 1060. To train a full StyleGAN 2 model for 256x256 images it takes about 3-5 days, but longer if you have more data and want to train longer. For my landscapes dataset I've been using a scraper to scrape them off the internet.
@@Matchue624 Thank you very much for your response. Please continue to make video tutorials and demonstrations, it is very inspiring)
Great, just wondering is it possible to run on a machine without GPU?
If I just want to inference.
Tutorial, please)