🗣️ AI Voice Cloner (XTTS v2)

A Streamlit web app for cloning voices using Coqui TTS.
Upload or record reference voices, fine-tune style & settings, and generate natural-sounding speech in 16+ languages — all through a clean, dark-themed interface.

---

🚀 Features

• 📁 Voice gallery management
  • Upload, record, save, activate, and delete reference voices
  • Batch delete multiple voices at once
  • Preview and switch between saved voices
• 🌍 Multi-language support
  • 16+ languages including English, Spanish, French, German, Chinese, Japanese, and more
• 🎭 Voice style controls
  • Choose between neutral, fast, and expressive styles
  • Optional speed and emotion overrides (happy, sad, angry, surprised, fearful)
• ⚙️ Advanced settings
  • Control voice stability and similarity for better cloning accuracy
  • Toggle loudness normalization for consistent output
• 📊 Generation history
  • Automatic metadata saving (text, style, reference voice, timestamp)
  • Recent generations gallery with preview, download, and re-synthesize option
• 💾 Multiple export formats
  • Download results as WAV or MP3 (requires FFmpeg)
• 🎨 Modern dark UI
  • Custom-styled Streamlit interface with smooth UX and mobile-friendly layout

---

🛠 Tech Stack

Frontend / UI

• Streamlit: Web interface for interaction and real-time updates
• streamlit-option-menu: Sidebar navigation with a clean UI
• Custom CSS styling for dark mode & better UX
• HTML5 MediaRecorder for in-browser recording

Backend / Logic

• Coqui TTS (XTTS v2): Voice cloning & text-to-speech model
• pydub: Audio file conversion (WebM/MP3 ↔ WAV)
• librosa & soundfile: Audio signal processing
• NumPy: Array and numerical operations

Utilities

• ffmpeg (system dependency): Required for audio encoding/decoding
• streamlit.components.v1 – custom recorder integration

---

📂 Project Structure

voice_clone_app/
│
├── outputs/                  # Generated outputs (auto-created)
│   └── xtts_*.wav            # Generated files
│
├── voices/                   # Reference voices (auto-created)
│   ├── ref_*.wav             # Uploaded/recorded samples
│
├── app.py                    # Main Streamlit application
├── run_app.py                # Launcher (cross-platform, auto-opens browser)
├── requirements.txt          # Python dependencies
├── LICENSE                   # Open-source license
├── .gitignore                # Ignored files/folders
└── README.md                 # Project documentation

---

⚙️ Installation

1. Clone the Repository

git clone https://github.com/EbrahimAR/AI-Voice-Cloner-XTTS-v2.git
cd AI-Voice-Cloner-XTTS-v2

2. Create virtual environment

python -m venv .venv
source .venv/bin/activate    # On Linux/Mac
.venv\Scripts\activate       # On Windows

3. Install dependencies

pip install -r requirements.txt

4. Install system dependencies

• ffmpeg is required for MP3 conversion:
  • Windows: choco install ffmpeg
  • macOS: brew install ffmpeg
  • Linux: sudo apt-get install ffmpeg

---

▶️ Run the app

streamlit run app.py

Or use the helper script:

python run_app.py

App will open at: http://localhost:8502

---

👨‍💻 Author

Ebrahim Abdul Raoof

LinkedIn | GitHub

---

📜 License

This project is licensed under the MIT License. See LICENSE for details.