Logo Lanfrica

Accarie/Kinyarwanda-assistant

Domaine:

natural language processing

Type de record:

software
Créateur:
Acc
Hôte:
Kinyarwanda Voice Assistant is a bilingual voice interaction tool that enables speech recognition and text-to-speech synthesis in Kinyarwanda and English. It uses OpenAI Whisper for automatic speech transcription, Coqui TTS for speech generation, and features an intuitive Gradio web interface for voice-based Q&A and custom speech synthesis. # Kinyarwanda-assistant Kinyarwanda Voice Assistant is a bilingual voice interaction tool that enables speech recognition and text-to-speech synthesis in Kinyarwanda and English. It uses OpenAI Whisper for automatic speech transcription, Coqui TTS for speech generation, and features an intuitive Gradio web interface for voice-based Q&A and custom speech synthesis. # 🗣️ Kinyarwanda Voice Assistant This project is a bilingual voice assistant that leverages Whisper for speech recognition and TTS for converting text into speech. It features a user-friendly Gradio interface with two main functions: voice-based question answering and custom text-to-speech generation. --- ## 🔧 Core Features - 🎙️ **Speech Recognition** via OpenAI Whisper. - 💬 **Preset Kinyarwanda Q&A** matching based on transcribed input. - 🔊 **Text-to-Speech Output** using Hugging Face-supported TTS models. - 🖥️ **Web-based Interface** created with Gradio. - 🔁 **Voice Cloning Support** using uploaded reference audio. - 🌐 **Deployable Interface** with optional public sharing. --- ## 🛠️ Setup Instructions ### ⚙️ Install Python Libraries First, install the necessary Python packages: ```bash pip install -q openai-whisper pip install numpy==1.24.3 --force-reinstall pip install gradio pip install transformers pip install torchaudio pip install TTS pip install nemo-toolkit