#122 · Primary category: Image Generation

ArtLine

ai art-portraits colab creative-coding deep-learning generative-art machine-learning movie-poster nogan nogan-training portraits

A Deep Learning based project for creating line art portraits.

Project last updated:03/03/23

GitHub Stars

3.6K

Forks

373

Contributors

2

License

MIT

Why we included this project

ArtLine turns ordinary portrait photos into stylized line art using a lightweight approach that skips the usual GAN critic entirely. For artists and designers who want clean facial line drawings without installing anything, the ready-made Colab notebooks are the fastest way in, and you can try both the smooth and quality model variants. The project also pairs the model with ControlNet, so you can guide the line-art style with written instructions instead of just feeding in a source image. The training details are documented openly, covering progressive resizing, perceptual loss on a pretrained U-Net, and a combined APDrawing and anime sketch dataset, which makes it a useful reference for anyone curious about facial line extraction without adversarial training.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category