🗣️ Voice Activity Detection with WebRTC & Silero

A cross-model Voice Activity Detection (VAD) tool with real-time and file-based analysis using both WebRTC VAD and Silero VAD models, built in PyQt6. Designed for visualization, evaluation, and comparison.

---

🚀 Features

• 🎙️ Live microphone-based VAD
• 📂 File-based VAD analysis
• 📊 Comparison mode: WebRTC vs Silero metrics side by side
• 📈 Generates:
  • Spectrograms
  • Waveforms
  • Confusion matrices
  • Metric charts (Accuracy, Precision, Recall, F1-score, etc.)
• 🧠 Silero model with intelligent frame-by-frame classification
• 🌐 WebRTC model integrated for fast binary VAD

---

🧰 Tech Stack

• Python 3.9+
• PyQt6
• matplotlib, numpy
• simpleaudio
• wave
• WebRTC VAD wrapper
• Silero VAD

---

💻 Installation

1. Clone the repo

git clone https://github.com/edyamza/Voice-Activity-Detection-WebRTC-Silero.git
cd Voice-Activity-Detection-WebRTC-Silero

2. Install dependencies

pip install -r requirements.txt

3. Run the app

python vad_guide.py

---

🖼️ GUI Preview

🎛️ Main Interface

📊 Confusion Matrices – WebRTC vs Silero

📈 Metric Comparison Bar Chart

🔊 Spectrogram + Waveform View

---

📁 Project Structure

├── vad_guide.py            # Main GUI application
├── vad_rec.py              # WebRTC file-based VAD
├── vad_rec_silero.py       # Silero VAD interface
├── vad_live.py             # Live audio processing
├── evaluation.py           # Metrics & graph generation
├── output/                 # Saved plots and images
└── requirements.txt

---

📄 License

This project is licensed under the MIT License.

---

✨ Author

Eduard Amza — GitHub

---

🧠 Inspired by

• Silero VAD
• WebRTC VAD

---

Feel free to ⭐ the project or contribute!