Anime2Sketch Hands-On

The world of machine learning is powering a whole new generation of interesting tools and the Anime2Sketch is yet another. Anime2Sketch is an open source MIT licenses project hosted on GitHub that enables you to take anime images and turn them in to sketches. Anime2Sketch is based on the Cornell University paper “Adversarial Open Domain Adaption for Sketch-to-Photo Synthesis” which is itself open source under the CC-By creative commons license and can be downloaded in PDF form here.

Anime2Sketch lists the following requirements:

That said, I was able to run it without issues using Windows, so the OS requirements seem to be inaccurate. Python is quickly becoming the language of choice for machine learning based projects and PyTorch is the library doing the heavy lifting. Installation is fairly simple assuming you have Python 3 and PIP installed, clone the repository, run the command to install the depencies then download the weights from the Google Drive folder and put it in the appropriate directory.

You can see Anime2Sketch in action in the video below. You can see several of the results of using Anime2Sketch in the repositories demos folder.

Scroll to Top