Suraksha - The Safety App

A smart safety app that detects male and female voices in an audio recording, classifies them, and counts unique voices. If an alert is triggered (e.g., an unusual male-to-female ratio or detected threats), it sends an SMS to trusted contacts. Additional features include real-time gender classification, location-based alerts, and analysis of risk zones. The app aims to enhance safety by enabling proactive alerts and helping users feel secure in various environments.

Inspiration

Suraksha addresses the critical issue of women’s safety by providing a proactive solution to detect and analyze potential threats. It uses real-time audio classification to identify male and female voices, count unique voices, and monitor surroundings. In high-risk situations, the app sends alerts to trusted contacts, shares live location data, and notifies users of nearby threats or high-risk zones. This solution helps prevent attacks, supports swift intervention, and empowers women with the tools needed to stay safe in various environments.

Features

• Gesture Analysis by Voice Command or Pattern:
  Trigger SOS alerts and share live location with trusted contacts and local police using voice commands on the screen.

• Real-Time Gender Classification:
  Utilize real-time audio capture to classify gender through analysis of facial features, body structure, and voice.

• Gender Distribution from Video and Audio:
  Analyze gender distribution in public spaces using both voice data to monitor and count the number of men and women present.

• Lone Woman Alert & Get Home Safe:
  Trigger alerts when a woman is detected alone at night with no nearby users. Her live location is shared with trusted contacts at 10-minute intervals.

• Nearby Men Detection & Alerts:
  Analyze the proximity and number of men around a woman by checking the type of nearest users' location and alerting women of potential threats.

• Risk Zone Classification & Alerts:
  Identify and notify users of high-risk areas (red zones) based on the analysis of previous incidents and alerts.

• Local Helplines & Emergency Contacts:
  Contact nearby police, hospitals, and pharmacies using GPS location and a list of emergency contacts for quick communication.

• Alert Nearby Users:
  When the Alert button is triggered, an alert notification is sent to nearby app users along with the user’s current location.

Tech Stack

• Framework : Flutter
• Language: Dart, Python
• Cloud services: Firebase
• Developer Tools: VS Code, Android Studio
• Version control: Git
• ML Library Used: TensorFlow
• Backend: Node.js (ExpressJs)
• SMS Service : Fast2SMS API

Installation & Setup

Clone the repository:

git clone https://github.com/Sohampal001/Women-Safety-Team-Explorer.git

Backend Setup

To set up the backend:

1. Navigate to the backend folder:

   cd Women-Safety-Team-Explorer/backend

2. Install dependencies: If using Node.js, run:

   npm install

3. Set up environment variables:
   Create a .env file and add necessary variables.

   PORT=5000
   YOUR_FAST2SMS_API_KEY=your_fast2sms_api_key
   FIREBASE_API_KEY=your_firebase_api_key
   FIREBASE_AUTH_DOMAIN=your_firebase_auth_domain
   FIREBASE_PROJECT_ID=your_firebase_project_id
   FIREBASE_STORAGE_BUCKET=your_firebase_storage_bucket
   FIREBASE_M
   

4. Run the backend:

   npm run dev

Frontend Setup

• Navigate to the Flutter directory and run:

  flutter pub get
  flutter run

API Endpoints

Method	Endpoint	Description
POST	/api/predict	Uploads audio and returns gender classification results, and sends SMS alerts to trusted contacts.

Contributing

1. Fork the repo
2. Create a new branch (feature-xyz)
3. Commit your changes
4. Open a pull request