cielo-b/Kinyarwanda-Voice-Assistant

# Kinyarwanda Voice Assistant 🤖🗣️ **Developer**: Cielo B. (cielo-b) **Email**: irumvaregisdmc@gmail.com **Course**: Intelligent Robotics **Instructor**: Gabriel Baziramwabo **Institution**: Rwanda Coding Academy ## 📌 Table of Contents - Project Overview - Key Features - Detailed Functionality - System Requirements - Installation Guide - Usage Instructions - Project Structure - Technical Stack - License - Contact ## 🌍 Project Overview This Kinyarwanda Voice Assistant simulates how humanoid robots process voice interactions in local languages. The system: 1. **Listens** to Kinyarwanda speech (like robot ears) 2. **Understands** the question (like robot brain) 3. **Responds** with spoken answers (like robot mouth) Developed for the Intelligent Robotics course, it demonstrates how robots can serve Rwandan communities in their native language. ## ✨ Key Features | Feature | Technology Used | Purpose | |---------|-----------------|---------| | Speech-to-Text | KinyaWhisper (ASR) | Convert Kinyarwanda audio to text | | Question Matching | NLP Dictionary | Map questions to correct answers | | Text-to-Speech | gTTS | Generate Kinyarwanda voice responses | ## 🔍 Detailed Functionality ### 1. Speech Recognition (ASR) - Uses KinyaWhisper (a Whisper model fine-tuned for Kinyarwanda) - Processes audio files or microphone input - Handles different audio formats (MP3, WAV) - Outputs accurate Kinyarwanda transcriptions ### 2. Question Understanding (NLP) - Pre-defined dictionary of 5+ QA pairs - Exact matching algorithm (expandable to fuzzy matching) - Returns "I don't know" for unmatched questions ### 3. Voice Response (TTS) - Google's Text-to-Speech (gTTS) with Kinyarwanda support - Generates clear, natural-sounding speech - Saves responses as MP3 files - Auto-plays responses in the interface ## 💻 System Requirements - Operating System: Windows 10+, macOS 10.15+, or Linux - Python: 3.8 or higher - RAM: Minimum 4GB (8GB recommended) - Disk Space: 2GB+ for models and …