tamarott/SinGAN
SinGAN is a PyTorch implementation of a GAN-based generative model that learns to produce diverse samples from a single natural image.
Not currently ranked — collecting fresh signals.
star history
SinGAN trains a generative adversarial network on a single natural image, learning to synthesize random samples, animation sequences, and perform image manipulation tasks like harmonization and super-resolution. The model learns internal statistics of the image at multiple scales to generate realistic variations. This approach won the Marr Prize at ICCV 2019 as best paper.
Frequently asked
- What is tamarott/SinGAN?
- SinGAN is a PyTorch implementation of a GAN-based generative model that learns to produce diverse samples from a single natural image.
- Is SinGAN open source?
- Yes — tamarott/SinGAN is an open-source project tracked on heatdrop.
- What language is SinGAN written in?
- tamarott/SinGAN is primarily written in Python.
- How popular is SinGAN?
- tamarott/SinGAN has 3.3k stars on GitHub.
- Where can I find SinGAN?
- tamarott/SinGAN is on GitHub at https://github.com/tamarott/SinGAN.