High-Performance Gait Recognition System with Firebase Integration

A real-time gait recognition application built with Streamlit, PyTorch, and Firebase, featuring high-performance camera processing, cloud authentication, and scalable data storage.

Features

• Real-time Gait Recognition: Continuous live camera feed with pose detection
• High-Performance Processing: GPU acceleration, batch processing, and optimized inference
• Firebase Authentication: Secure cloud-based user authentication and management
• Cloud Data Storage: Firestore integration for scalable gait pattern storage
• Gait Enrollment: Multi-sequence gait data collection with configurable parameters
• Data Management: CRUD operations, batch comparisons, and analytics
• User Profiles: Comprehensive user management and statistics
• Clean Interface: Professional UI without visual clutter

Requirements

• Python 3.10+
• CUDA-compatible GPU (recommended)
• Webcam
• Firebase project (see setup instructions below)

Firebase Setup

Before running the application, you need to set up Firebase:

1. Create Firebase Project

   • Go to Firebase Console
   • Create a new project
   • Enable Authentication (Email/Password)
   • Create Firestore Database

2. Configure Authentication

   • See FIREBASE_SETUP.md for detailed instructions
   • Set up Streamlit secrets or environment variables
   • Configure Firestore security rules

3. Install Firebase Dependencies

   pip install firebase-admin pyrebase4

Installation

1. Clone the repository

2. Create a virtual environment:

   python -m venv .venv
   .venv\Scripts\activate  # Windows

3. Install dependencies:

   pip install -r requirements.txt

4. Set up Firebase configuration (see FIREBASE_SETUP.md)

Usage

1. Start the application:

   streamlit run main.py

2. Access the application at http://localhost:8501

3. Create a Firebase account or login

4. Enroll gait patterns in the "Enroll Gait" tab

5. Use "High-Performance Recognition" for real-time identification

6. Manage your data in "My Data" tab

7. View your profile in "Profile" tab

Architecture

• Frontend: Streamlit web interface
• Authentication: Firebase Authentication with secure token management
• Data Storage: Firebase Firestore for cloud storage with local backup
• Models: YOLOv8n-pose for pose detection, FastPoseGait for gait recognition
• Cloud Integration: Real-time sync across devices and users

Performance Features

• GPU acceleration with CUDA
• Batch processing for improved throughput
• Threading for camera capture and processing
• Real-time streaming with minimal latency
• Data quality filtering (removes corrupted patterns)
• Firebase caching and offline support

File Structure

Recognition System/
├── main.py                           # Main application
├── firebase_setup/                   # Firebase configuration and services
│   ├── firebase_config.py            # Firebase configuration
│   ├── firebase_auth_service.py      # Firebase authentication
│   ├── firebase_user_manager.py      # User and data management
│   ├── FIREBASE_SETUP.md            # Firebase setup guide
│   └── gaitrecognitionapp-47c41-firebase-adminsdk-fbsvc-47efac1c13.json
├── models/                          # Gait recognition models
│   ├── gait_model.py                # OptimizedFastPoseGaitModel
│   └── __init__.py
├── data/                            # Data management
│   ├── gait_data_manager.py         # OptimizedGaitDataManager
│   └── __init__.py
├── processing/                      # Camera processing
│   ├── camera_processor.py          # HighPerformanceCameraProcessor
│   └── __init__.py
├── ui/                             # User interface
│   ├── render_functions.py          # UI rendering functions
│   └── __init__.py
├── utils/                          # Utility functions
│   ├── data_analysis.py             # Data analysis and visualization
│   └── __init__.py
├── local_auth/                     # Local authentication fallback
│   ├── simple_auth_working.py      # Local authentication system
│   ├── simple_data_manager.py      # Local data management
│   └── __init__.py                 # Package initialization
├── requirements.txt                # Dependencies
├── yolov8n-pose.pt                # YOLO pose detection model
├── gait_data/                     # User gait data storage
│   ├── demo_gait.pkl
│   └── Hashir_gait.pkl
└── README.md                       # This file

Firebase Integration Features

• Secure Authentication: Firebase Auth with email/password
• Cloud Storage: Firestore for gait patterns and user data
• Real-time Sync: Data synchronization across devices
• User Management: Profile management and statistics
• Scalability: Supports multiple users and large datasets
• Backup System: Local fallback for offline operation

Technical Details

• Pose Detection: YOLOv8n-pose model
• Gait Recognition: FastPoseGait neural network
• Feature Extraction: 128-dimensional gait features
• Recognition Threshold: 50% confidence
• Data Quality: Filters patterns with norms > 1000
• Cloud Storage: Firestore with automatic backup to local files

Migration from Local Auth

If you're upgrading from the local authentication system:

1. Backup existing data:

   • Export gait data from "My Data" tab
   • Save user_data.json file

2. Set up Firebase:

   • Follow FIREBASE_SETUP.md instructions
   • Create Firebase accounts for existing users

3. Import data:

   • Use the new Firebase-based enrollment system
   • Data will be automatically synced to the cloud

Troubleshooting

• Firebase Connection Issues: Check your configuration in FIREBASE_SETUP.md
• Authentication Errors: Verify Firebase project settings and API keys
• GPU Issues: Ensure CUDA drivers are installed for acceleration
• Webcam Issues: Check webcam permissions and availability
• Data Sync Issues: Check Firestore security rules and permissions

Security Features

• Firebase Security Rules: User-specific data access
• Token Management: Secure authentication tokens
• Data Encryption: Firestore automatic encryption
• Access Control: Role-based permissions

License

This project is for educational and research purposes.