Skip to content

Releases: rudyoactiv/typescribe-handwriting

TypeScribe - Text to Handwriting

14 Dec 11:27
@rudyoactiv rudyoactiv
v1.0
2afd04d
Compare
Choose a tag to compare
TypeScribe - Text to Handwriting Latest
Latest

TypeScribe v1.0

Need to send out a hundred organic Christmas cards? Or you have a fifty page essay due and Brush Script doesn't quite cut it?
Got you covered.

TypeScribe is a tool that uses Machine Learning to convert typed text into realistic handwriting by introducing customizable degrees of randomness and variations, to make it all so believable.

TypeScribe Totally not homework A truly heartfelt letter Christmas Cards too!?

Features

  • Realistic Handwriting Generation with a Recurrent Neural Network (RNN)
  • Choose from 12 predefined handwriting styles.
  • Adjust line spacing, page size, margins, ink color, pen thickness, and more.
  • User-friendly interface
  • Scalable SVG Output that maintains quality even on resizing

System Requirements

  • Operating System: Microsoft Windows.
  • Other Requirements: 7Zip or WinZip to extract the files.

Download

Download Windows Executable

Installation

  1. Download the TypeScribe-v1.0.zip file from the Assets section below.
  2. Extract the contents to a folder on your system.
  3. Run the TypeScribe.exe file to start the application.

How to Use

  1. Launch the application.
  2. Enter your text and adjust settings via the interface.
  3. Preview the layout and handwriting style.
  4. Click Generate.
  5. Select a destination folder for your files.
  6. Wait while it generates your output. May take a minute.

Known Issues

  • Your Antivirus may flag (and even delete) the executable. This is due to the way Python programs are built and handled by Windows and cannot possibly be fixed. However, there is no cause for concern. If you would rather run this program directly from the Python file, feel free to clone the repository.
  • The application itself may freeze while your file is being generated. No need to panic, it does not crash.

Acknowledgements

This project is inspired by the work in the handwriting-synthesis repository by sjvasquez, which provides the foundational implementation for handwriting synthesis using Recurrent Neural Networks (RNNs).

The handwriting synthesis in TypeScribe is based on the work presented in the paper Generating Sequences with Recurrent Neural Networks by Alex Graves. [Graves, A. (2013). Generating sequences with recurrent neural networks. arXiv preprint arXiv:1308.0850.]

Thank you for trying TypeScribe! ✍️

Assets 3
Loading