AI-Powered Decentralized Identity Verification System

Overview

DID++ is a multi-modal identity verification system combining:

• Face biometrics (MediaPipe)
• Voice biometrics (MFCC)
• Document verification (OCR)
• Weighted score fusion

Designed for Self-Sovereign Identity (SSI) with future blockchain integration.

Features

• Face recognition using MediaPipe landmarks
• Voice verification using MFCC embeddings
• Document OCR verification using EasyOCR
• Multi-modal score fusion
• Privacy-preserving design (no raw biometric storage)

Requirements

• Python 3.11.9

Installation

1. Install Python 3.11.9

   • Download Python 3.11.9 from the official Python website.
   • Follow the instructions for your operating system.
   • Verify the installation:

     python --version
     pip --version

     Expected output:

     Python 3.11.9
     

2. Install dependencies

   • It is recommended to use a virtual environment:

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

   • Install required packages:

     pip install -r requirements.txt

Running Tests & Generating Reports

To run all tests:

pytest tests/

To generate a PDF report for the LFW face verification test, set the environment variable LFW_PDF_REPORT=1 before running tests. You can do this in two ways:

Recommended: Create a .env file in the project root with:

LFW_PDF_REPORT=1

The project uses python-dotenv to automatically load this variable.

Or, set it in your shell:

• PowerShell:

  $env:LFW_PDF_REPORT=1; pytest tests/

• Command Prompt:

  set LFW_PDF_REPORT=1 && pytest tests/

The PDF report will be saved as lfw_face_verification_report.pdf in the project directory.

To see detailed output in the terminal, use:

pytest tests/ -s

The -s flag ensures all print output is shown.

Running the Application

To start the Flask app, set the following environment variables and use the Flask CLI:

Required:

• FLASK_APP=app.py
• FLASK_ENV=development (optional, enables debug mode)

On Windows (PowerShell):

$env:FLASK_APP="app.py"
flask run

On Windows (Command Prompt):

set FLASK_APP=app.py
flask run

The app will be available at http://127.0.0.1:5000/

Environment Variables

• LFW_PDF_REPORT=1 — Enables PDF report generation for LFW face verification tests.
• FLASK_APP=app.py — Required for flask run.
• FLASK_ENV=development — (Optional) Enables debug mode for Flask.

If using a .env file, all variables will be loaded automatically if python-dotenv is installed (already in requirements).