Autolock V4 - Face Recognition System

Overview

Autolock V4 is a face recognition system designed to enhance security by automatically locking and unlocking your computer based on user presence. The system utilizes a YOLOv8 model for face detection and K-Nearest Neighbors (KNN) for face recognition. It can manage a blacklist of faces and prevent unauthorized access.

Features

• Real-time face detection and recognition
• Automatic system locking when user is absent
• Blacklist management for unauthorized faces
• User-friendly interface for creating master profile
• Multi-platform support (Windows, macOS, Linux)

Version History

Autolock V1 (Initial Release)

• Basic face detection using YOLOv8
• Simple face recognition with KNN classifier
• Manual system lock when no face detected
• Basic webcam handling and UI display

Autolock V2 (Security Update)

• Added blacklist system for unauthorized faces
• Improved face recognition accuracy with normalized features
• Increased inactivity timeout to 10 seconds (from 5)
• Added validation during profile creation
• Enhanced error handling and logging

Autolock V3 (Auto-Unlock Update)

• Added automatic unlock feature using PIN
• Introduced system lock state tracking
• Added lock timestamp monitoring
• PIN-based security for unlock operations
• Improved recognition confidence handling
• Enhanced profile creation with real-time feedback

Autolock V4 (Current - Performance Update)

• Improved recognition speed and accuracy
• Added rolling confidence average for stability
• Reduced false positives with stricter thresholds
• Enhanced feature extraction with HOG
• Added auto-recovery from failed recognition
• Improved multi-threading for lock monitoring
• Auto-switching between available cameras

Project Structure

Core Files

• autolock_v4.py: Main application file containing the face recognition system
• requirements.txt: Lists all Python dependencies
• yolov8n-face.pt: YOLOv8 face detection model (download separately)

Configuration & Data

• face_data/: Directory storing face recognition data
  • master_embeddings.pkl: Stored face embeddings for master user
  • master_knn_model.pkl: Trained KNN model for face recognition
  • blacklist.pkl: List of blocked face embeddings
  • master_feature_dim.txt: Feature dimensions configuration

Testing & Setup Files

• tests/
  • test_face_recognition.py: Unit tests for face recognition system
  • test_system_lock.py: Tests for system locking functionality
  • test_blacklist.py: Tests for blacklist management
  • conftest.py: PyTest configuration and fixtures

Helper Scripts

• scripts/
  • setup_camera.py: Utility to test and configure webcam
  • model_download.py: Script to download YOLOv8 face model
  • profile_cleanup.py: Tool to reset master profile
  • performance_test.py: Benchmarking script

Documentation

• docs/
  • setup_guide.md: Detailed installation instructions
  • troubleshooting.md: Common issues and solutions
  • api_reference.md: API documentation
  • CONTRIBUTING.md: Guidelines for contributors

Testing

Run the test suite:

python -m pytest tests/

Run specific test categories:

python -m pytest tests/test_face_recognition.py
python -m pytest tests/test_system_lock.py

Generate test coverage report:

python -m pytest --cov=src tests/

Development Setup

1. Clone and install dependencies:

git clone <repository-url>
cd autolock-v4
pip install -r requirements.txt

2. Download YOLOv8 model:

python scripts/model_download.py

3. Test camera setup:

python scripts/setup_camera.py

4. Run performance tests:

python scripts/performance_test.py

Installation

Prerequisites

• Python 3.7 or higher
• pip (Python package installer)

Steps

1. Clone the repository:

   git clone <repository-url>
   cd autolock-v4
   

2. Install the required packages:

   pip install -r requirements.txt
   

Usage

To run the face recognition system:

python src/autolock_v4.py

Optional arguments:

• --model: Path to YOLOv8 face detection model
• --data-dir: Directory for face recognition data
• --master: Name for master user
• --timeout: Inactivity timeout before locking
• --tolerance: Recognition confidence tolerance
• --retrain: Force retraining of master profile

License

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

Acknowledgments

• YOLOv8 for face detection
• OpenCV for image processing
• scikit-learn for machine learning algorithms

For any issues or feature requests, please open an issue on the GitHub repository.