Viateur-akimana/KinyaVoice

A lightweight Kinyarwanda speech assistant that transcribes spoken queries, processes natural language intents, and responds with synthesized speech—bridging technology with Rwanda's primary language. # Kinyarwanda Voice Assistant A mini voice assistant for Kinyarwanda language, built for the Intelligent Robotics course (April 2025). This project implements a complete voice interaction pipeline with Automatic Speech Recognition (ASR), Natural Language Processing (NLP), and Text-to-Speech (TTS) capabilities. ## 🚀 Features - **🗣️ Speech Recognition**: Transcribes Kinyarwanda speech to text using `benax-rw/KinyaWhisper` (ASR) - **🤖 NLP Processing**: Understands queries using intent-based pattern matching with fuzzy matching capabilities - **🔊 Speech Synthesis**: Generates natural-sounding Kinyarwanda speech using custom TTS models - **🌐 Web Interface**: User-friendly Gradio interface for interactive conversations - **💻 CLI Support**: Command-line interface for batch processing of audio files ## 📂 Project Structure ``` kinyarwanda-voice-assistant/ ├── gradio_app.py # Main Gradio web interface ├── main.py # Command-line application ├── asr/ │ └── transcribe.py # Speech recognition module ├── nlp/ │ ├── qa_engine.py # Question answering system │ └── intents.json # Intent patterns and responses ├── tts/ │ └── speech_synth.py # Text-to-speech engine ├── Inference/ │ └── kinyatts/ # TTS model and utilities ├── test_audio/ # Sample audio files └── transcription_output/ # Generated audio outputs ``` ## Installation 1. Clone the repository: ```bash git clone github.com cd kinyarwanda-voice-assistant ``` 2. Create and activate a virtual environment: ```bash python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate ``` 3. Install dependencies: ```bash pip install -r requirements.txt ``` 4. Set up the TTS model: - Download the TTS model file and place it at the correct location - Verify the paths in `tts/speech_synth.py` ## Usage You can use the system from the command line: ```bash python main.py ` …