Turning images into coloring book pages with infinite resolution.
The following features are supported:
Binaries are available in the Releases section.
To get started with the project, do the following steps:
make init
to download Meta's Segment Anything model and create a Python venv with all required dependencies.make build
to start the full build pipeline. This runs Pyinstaller on the ML model and then runs Electron Forge to create the final packaged application. The only currently supported format is a zip
file, but other formats are supported by Electron Forge.The frontend is an Electron application built with the latest web technologies, including React, TypeScript, and Tailwind CSS. A context bridge is used so that the renderer can send calls to the Node.js backend which has access to system processes.
The backend is a full ML pipeline that takes in a full color image and outputs a sketch in vector format. The pipeline is outlined below:
Pyinstaller is used to compile the Python scripts into binary executables. These binaries are then copied over to the resources
folder of the Electron app during the build process of Electron Forge.