Skip to content

Introduction

Welcome to the documentation for the Igbo Voice Assistant, a project dedicated to building a voice-based personal assistant that understands and speaks the Igbo language.

This project aims to create a simple, proof-of-concept voice assistant that can be run locally. It's a learning project, and the documentation will cover all phases of the project, from data collection and model training to application development.

High-Level Overview

The system is composed of a frontend web application and a backend processing server. The user interacts with the frontend to record their voice. The frontend then sends this audio to the backend, which performs a series of transformations to generate a spoken response in Igbo.

The entire process is designed to be a seamless pipeline, transforming raw audio input into a final audio output.

System Flowchart

UserFrontendBackendSpeech-to-Text (Igbo)Translate to EnglishGenerate Response (English)Translate to IgboText-to-Speech (Igbo) Records voice Sends audio data (POST /voice) Returns audio URL (GET /storage/{file_id}.wav) Plays audio response