Introduction

Augmentative and Alternative Communcation (AAC) refers to any tools, strategies, or devices that help people communicate to others when they have difficulties speaking. They can include simple tools like picture boards to high tech devies like a tablet or an app.

An AAC Board, which is what the project is primarily targeting as the main device for our API, is a visual interface with images, words, or symbols that can represent basic phrases or ideas, which the user can select items on a board, which is then spoken out aloud by the device.

Project Overview

This project implements a Speech-to-Text API designed for use with AAC systmes. The API receives an audio file (such as a spoken message or sound recording), processes it through a Python-base speech recognizer, and returns a JSON response contaning the recognized text, metadata, and error information in order for game developers developing AAC based games to map the returned response to a game input

Project Goals

• Provide a lightweight, open-source speech-to-text API for AAC use cases.
• Support multi-format audio inputs (WAV, FLAC, AIFF, MP3, etc.).
• Implement dynamic background noise cancellation for accessibility and reliability for AAC Board recording
• Enable easy integration into AAC games and apps to assist users with speech limitations and to assist game developers to include speech functionanlity in their games.

This documentation serves as an extensive description of the project and API in order for any end-user or developer to be able to understand and fully utilize all tools and libraries that the API provides.