Categories
Ongoing Projects

Tafsiri S2ST

Tafsiri Logo

Project Overview

ProjectTafsiri
Lead PersonCollins Mwange
ASL/OC4ESASL

Introduction

Tafsiri is an AI-powered software that provides speech-to-speech translation (S2ST) features. The application is designed to transcribe and translate audio and output text and audio in the target language.

huggingface.co

Tafsiri uses two LLMs to achieve S2ST:

  • SeamlessM4T v2 – Meta’s seamlessM4T handles speech-to-text (S2T) translation. e.g. speech in language X to text in language Y.
  • Edge TTS – Microsoft’s edge TTS handles text-to-speech (T2S) translation. The model takes as input the output from seamlessM4T. e.g. text in language Y => speech in language Y.

User Requirements Gathering

To learn more and suggest/request features, go here: https://creativedisturbance.org/index.php/tafsiri/

This page will be updated from time to time. Stay tuned!