Skip to content

Home

Jump to bottom
Walker Herring edited this page May 23, 2021 · 12 revisions

Technical Overview

  • In order to determine which commands are most common, we analyzed thousands of documents with the Top LaTeX Commands module
  • LaTeX is parsed by the TexParser module
  • Math mode content is made ready for text-to-speech by the MathModeToSSML module
  • The configuration system allows the user to configure how certain aspects of the documents are spoken. Developers can also set defaults.
  • To ensure audio is comprehensible, we mark up text with SSML before text-to-speech generation.
  • Once LaTeX has been converted to human-readable text and marked up with SSML, AWS Polly is used to generate spoken audio.

Opportunities for future development can be found in the wiki pages under Future Work. Please read these sections before starting your own development on this project.

BrianHutchinson Research 2021

Clone this wiki locally