microsoft/StyleSwin
A Transformer-based GAN architecture for high-resolution image synthesis, achieving state-of-the-art results on CelebA-HQ and FFHQ benchmarks.
Not currently ranked — collecting fresh signals.
star history
This repository provides the official PyTorch implementation of the StyleSwin paper, a generative adversarial network that replaces convolutional layers in StyleGAN with Swin Transformer blocks. The architecture achieves high-resolution image generation by leveraging the local attention mechanism of Swin transformers while maintaining the style-mixing capabilities of StyleGAN. It supports generation at multiple resolutions including 256x256 and 1024x1024.
Frequently asked
- What is microsoft/StyleSwin?
- A Transformer-based GAN architecture for high-resolution image synthesis, achieving state-of-the-art results on CelebA-HQ and FFHQ benchmarks.
- Is StyleSwin open source?
- Yes — microsoft/StyleSwin is open source, released under the MIT license.
- What language is StyleSwin written in?
- microsoft/StyleSwin is primarily written in Python.
- How popular is StyleSwin?
- microsoft/StyleSwin has 544 stars on GitHub.
- Where can I find StyleSwin?
- microsoft/StyleSwin is on GitHub at https://github.com/microsoft/StyleSwin.