Voice Verification System

🎙️ A Simple Yet Powerful Voice Identity Verification Application

This application, built using React and FastAPI, provides an intuitive and efficient way to register and verify user identities through voice comparison.

Frontend:

Backend:

Database:

Programming Languages:

Table of Contents

• Screenshots
• Introduction
• Key Features
• Architecture
  • Frontend Structure
  • Backend Structure
• Installation
• Usage
• API Endpoints
• Known Issues and Limitations
• Contributing
• License
• Acknowledgments

Screenshots

Figure 1: Adding a new user with voice sample

Figure 2: Voice comparison interface

Figure 3: User management interface

Figure 4: Deleting a User

Figure 5: Shimmer loading animation

Introduction

The Voice Verification System is a modern web application that allows users to register their voice and later verify their identity through voice comparison. It combines the power of React for the frontend and FastAPI for the backend to create a seamless and efficient user experience.

Key Features

🎤 Voice Registration

• Users can register by providing personal information and a voice sample.
• Secure storage of user data and voice embeddings.
• Form validation ensures accurate and complete user information.

🔍 Voice Comparison

• Compare a new voice sample against registered users.
• Receive similarity scores and verification results.
• Dimensional visualization of voice embeddings for intuitive comparison.

👥 User Management

• Add new users through an intuitive form that has validation.
• View and manage registered users through an intuitive interface.
• Delete user profiles when necessary.

📊 Data Visualization

• Interactive radar chart displaying multi-dimensional voice embeddings.
• Adjustable number of dimensions for detailed analysis.

⏳ Shimmer Loading

• Smooth loading animations during data fetching and processing.
• Enhances user experience by providing visual feedback on loading states.

🖥️ Responsive UI

• Modern, responsive design using PrimeReact components.
• Intuitive navigation between different sections of the app.

⚡ Fast and Efficient

• Utilizes Vite for rapid development and optimized builds.
• FastAPI backend for high-performance API operations.

🔊 Advanced Audio Processing

• Leverages PyAnnote for sophisticated audio analysis and feature extraction.

🗃️ Efficient Data Storage

• Uses ChromaDB for vector storage, enabling fast similarity searches.
• SQLite database for structured data storage.

Architecture

Frontend Structure

The React frontend is organized as follows:

voice-verification/
├── public/
├── src/
│   ├── components/
│   │   └── Navbar.css
│   │   └── VoiceComparisonChart.jsx
│   ├── pages/
│   │   ├── AddUserPage.jsx
│   │   └── ComparePage.jsx
│   │   └── UsersPage.jsx
│   │   └── NotFoundPage.jsx
│   ├── App.css
│   ├── App.jsx
│   └── main.jsx
├── index.html
├── package.json
├── package-lock.json
└── vite.config.js

Backend Structure

The FastAPI backend follows this structure:

backend/
├── app/
│   ├── api.py
│   ├── models/
│   │   ├── custom_exceptions.py
│   │   ├── user.py
│   │   └── pydantic_models.py
│   ├── repositories/
│   │   ├── user_repository.py
│   │   ├── chroma_repository.py
│   │   └── audio_repository.py
├── main.py
├── requirements.txt

Installation

To install and run the Voice Verification System, follow these steps:

1. Open two terminal windows.

2. In the first terminal:
   a. Create and activate a virtual environment (recommended): Using Python's built-in venv:

   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

   Or using Conda:

   conda create -n voice_verification python=3.9
   conda activate voice_verification

   b. Navigate to the backend directory:

   cd backend

   c. Install the Python requirements:

   pip install -r requirements.txt

   d. Run the backend server:

   python main.py

3. In the second terminal:
   a. Navigate to the voice-verification folder:

   cd voice-verification

   b. Install the Node.js dependencies:

   npm install

   c. Start the development server:

   npm run dev

4. Open your web browser and go to the provided link, typically:

   http://localhost:5173
   

You should now have both the backend and frontend running, and you can access the Voice Verification System through your web browser.

Usage

1. Add User: Navigate to the "Add User" page to register a new user with their voice sample.
2. Compare Voice: Use the "Compare" page to record a voice sample and compare it against registered users.
3. Manage Users: The "Users" page allows you to view and manage registered users.

API Endpoints

Our Voice Verification System provides a comprehensive set of RESTful API endpoints for user management and audio processing. Below is a detailed overview of the available endpoints:

Method	Endpoint	Description
POST	/users	Register a new user with voice sample
GET	/users/{user_id}	Retrieve a specific user's information
PUT	/users/{user_id}	Update a user's information
DELETE	/users/{user_id}	Delete a user
GET	/users	Retrieve all registered users
GET	/users_with_embeddings	Retrieve all users with their voice embeddings
POST	/audio/process	Process and save a new voice sample for a user
POST	/audio/compare	Compare a voice sample against a registered user

Detailed Endpoint Descriptions

Register a New User

POST /users

• Body: multipart/form-data
  • name: string
  • surname: string
  • email: string
  • audio: file (voice sample)
• Response: User object with ID

Retrieve a Specific User

GET /users/{user_id}

• Parameters:
  • user_id: integer (path parameter)
• Response: User object

Update a User

PUT /users/{user_id}

• Parameters:
  • user_id: integer (path parameter)
• Body: JSON
  • name: string
  • surname: string
  • email: string
• Response: Updated user object

Delete a User

DELETE /users/{user_id}

• Parameters:
  • user_id: integer (path parameter)
• Response: Success message

Retrieve All Users

GET /users

• Response: Array of user objects

Retrieve Users with Embeddings

GET /users_with_embeddings

• Response: Array of user objects including voice embeddings

Process Audio

POST /audio/process

• Body: multipart/form-data
  • user_id: integer
  • file: file (audio sample)
• Response: Success message

Compare Voice Sample

POST /audio/compare

• Body: multipart/form-data
  • user_id: integer
  • file: file (voice sample to compare)
• Response: Similarity score and voice embeddings

Known Issues and Limitations

Known issues and limitations of the project include:

1. Email Uniqueness: While the system requires a unique email address for each user, it does not provide specific feedback when attempting to register with an already-used email. Instead, it returns a generic "Failed to register user" error, which may confuse users trying to sign up with an email that's already in the system.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• PrimeReact for UI components

• PyAnnote for audio processing

• ChromaDB for vector storage

• PyAnnote for audio processing

• ChromaDB for vector storage