Skip to content

This repository houses the development of a cutting-edge AI-powered Text-to-Speech (TTS) model tailored for the Hausa language. The project aims to create a state-of-the-art TTS system through iterative development, rigorous testing, and collaboration among machine learning engineers, data scientists, linguists, and software developers.

License

Notifications You must be signed in to change notification settings

AIBauchi/AI-Powered-Hausa-Text-to-Speech-Model

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 

Repository files navigation

AI-Powered Hausa Text-to-Speech Model.

Welcome! This document serves as a guide for the AI-Powered Hausa Text-to-Speech Model project, developed in collaboration with the AI Bauchi Community. It outlines the project goals, current progress, upcoming milestones, and communication channels.

Project Goal:

Develop a high-quality, AI-powered Text-to-Speech (TTS) model for the Hausa language, promoting its preservation and wider accessibility through natural-sounding voice technology.

Project Phases:

  • Data Collection and Preprocessing (Ongoing):

    • Objective: Gather a diverse and representative dataset of Hausa speech, ensuring cultural and linguistic nuances are considered.
    • Progress: Currently collecting over 2000 speech samples from various speakers, dialects, and speaking styles. Data cleaning and pre-processing ongoing.
    • Checkpoint: Review data diversity and quality with AI Bauchi Community Leader by January 15th.'
  • Model Architecture and Training (Planning):

    • Objective: Choose and train a suitable TTS model architecture based on collected data and desired voice characteristics (e.g., female voice model named Amina).
    • Plans: Explore fine-tuning existing Coqui AI Hausa models and potentially training a custom model with additional data.
    • Team Meeting: Discuss model architecture options with AI Bauchi Community Leader on January 20th.
  • Testing and Evaluation (Future):

    • Objective: Assess the performance and naturalness of the trained model on various metrics.
    • Plans: Develop testing criteria and evaluation methods in collaboration with the AI Bauchi Community.
    • Checkpoint: Present preliminary model performance results to AI Bauchi Community on TBD .
  • Deployment and Scaling (Future):

    • Objective: Make the TTS model accessible for real-world applications and explore potential scaling options.
    • Plans: Investigate deployment platforms and consider community feedback for accessibility and usage.
    • Team Meeting: Discuss deployment strategies and user feedback with AI Bauchi Community on TBD.

Team Roles:

  • Project Lead: Amina Shiga
  • Linguistic Expert: Amina Gadiya.
  • AI Bauchi Community Leader: Nathaniel Handan
  • Additional roles to be filled based on project needs

Communication and Collaboration:

  • Weekly Progress Meetings: Every Thurdays at 04:00 PM WAT (online platform to be determined)
  • Team Meetings with AI Bauchi Community Leader: As outlined in project phases
  • Project Documentation: Maintain updated README and project wiki
  • Open Communication: Encourage active participation and feedback from all team members and the AI Bauchi Community

Next Steps:

  • Continue data collection and pre-processing, ensuring data diversity and quality.
  • Research and finalize model architecture options for discussion with the AI Bauchi Community Leader.
  • Develop testing criteria and evaluation methods for the trained model.
  • Set up communication channels and schedule regular meetings with the AI Bauchi Community Leader.

We believe this project has the potential to significantly contribute to the preservation and accessibility of the Hausa language. We invite everyone to participate and share their valuable insights!

Thank you for your support!

About

This repository houses the development of a cutting-edge AI-powered Text-to-Speech (TTS) model tailored for the Hausa language. The project aims to create a state-of-the-art TTS system through iterative development, rigorous testing, and collaboration among machine learning engineers, data scientists, linguists, and software developers.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published