VaaniRakshak is an intelligent, offline voice assistant tailored for disaster and emergency response scenarios. Designed to operate without internet connectivity, it provides real-time assistance using voice commands, helping users locate shelters, access emergency services, and receive critical information — even in infrastructure-compromised environments.
- 🔴 Emergency-themed interface
- 🎙️ Start/Stop voice control
- ✍️ Manual text input
- 🧭 Real-time updates and response messages
🔗 Watch the video demo on Google Drive
✅ Watch the assistant recognize phrases like “I need help”, “medical emergency”, and reply with shelter info, alerts, and voice feedback.
- 🎙️ Offline Speech Recognition using Vosk
- 🗣️ Text-to-Speech Synthesis with
pyttsx3 - 🖥️ User-Friendly GUI powered by Tkinter
- 📡 Emergency Response Engine with shelter lookup and SMS alert simulation
- 🌍 Location-aware guidance (mocked for offline support)
- 🔧 Fully modular and configurable architecture
“In disaster scenarios where internet access is disrupted, VaaniRakshak enables users to interact with emergency services and systems hands-free, offline, and in real-time — just by speaking.”
+---------------------+
| GUI Interface |
+----------+----------+
↓
+----------v----------+
| Voice Assistant | <-- Orchestrator
+----------+----------+
↓
+----------v----------+
| Speech Processor | <-- STT + TTS
+----------+----------+
↓
+----------v----------+
| Emergency Handler | <-- Logic + Lookup
+----------+----------+
↓
+----------v----------+
| Configuration Layer |
+---------------------+
| Layer | Tool/Library |
|---|---|
| Programming Lang | Python 3.11 |
| Speech Recognition | Vosk (offline) |
| Text-to-Speech | pyttsx3 |
| Audio Input/Output | PyAudio |
| GUI | Tkinter (standard Python GUI) |
| Data Storage | JSON |
VaaniRakshak/
├── main.py # Entry point
├── voice_assistant.py # Core controller
├── speech_processor.py # STT + TTS
├── emergency_handler.py # Logic/Response
├── gui_interface.py # Tkinter GUI
├── config.py # Configuration
├── data/
│ ├── emergency_phrases.json
│ ├── shelters.json
│ └── locations.json
├── models/
│ └── vosk-model-small-en-us-0.15/
└── README.md
git clone https://github.com/your-username/VaaniRakshak.git
cd VaaniRakshakpip install vosk pyttsx3 pyaudio💡 If
pyaudioinstallation fails:
pip install pipwin
pipwin install pyaudioDownload vosk-model-small-en-us-0.15 and extract it here:
VaaniRakshak/
└── models/
└── vosk-model-small-en-us-0.15/
├── am/
├── conf/
├── model.conf
└── ...
python main.py"I need help""Nearest shelter""Medical emergency""Fire alert""Exit"
Install PyInstaller:
pip install pyinstallerThen build:
pyinstaller --noconfirm --onefile --windowed main.pyPlace models/ and data/ folders alongside the .exe.
This project is licensed under the MIT License.
Pull requests are welcome! If you'd like to contribute improvements, localization, or accessibility features, please open an issue.
- 2025-06-14 – Initial release with full offline assistant functionality
Subuhana B
Developer | Tech for Social Good | Voice AI Enthusiast
email: [email protected]
Github: subuhana2303
“VaaniRakshak speaks for you — when your voice is the only tool you have.”