A comprehensive, modular face recognition framework
Demo • Features • Installation • Quick Start • Documentation • API
FR-Framework is a powerful, modular toolkit for face recognition tasks, offering everything from basic face detection to advanced facial analytics. Built with flexibility and performance in mind, it provides a complete solution for developers, researchers, and enterprises working with facial recognition technology.
- Modular Architecture: Use only the components you need
- Multiple Detection Models: Support for both HOG (CPU) and CNN (GPU) detection models
- Facial Landmark Detection: Identify and track 68 facial landmarks
- Face Analysis: Blur detection, head pose estimation, eye aspect ratio analysis
- Real-time Processing: Live camera feed integration
- Video Processing: Analyze video files with detailed timeline results
- REST API: Easy integration with other applications via a comprehensive API
- Docker Support: Simple deployment in containers
- Extensive Documentation: Tutorials, API references, and examples
 Python |
 OpenCV |
 dlib |
 FastAPI |
 Docker |
| Model | Speed (ms/image) | Accuracy | Platform |
|---|---|---|---|
| HOG (CPU) | 31.2 | 97.1% | Core i7, 16GB RAM |
| CNN (CPU) | 121.5 | 99.3% | Core i7, 16GB RAM |
| CNN (GPU) | 8.7 | 99.3% | NVIDIA RTX 3080 |
*Benchmarks based on 1080p images with varying numbers of faces
| Module | Description | Key Features |
|---|---|---|
| fr-photo | Photo-based face detection and recognition |
• Multiple detection models (HOG/CNN) • Face embedding generation • Recognition against known faces |
| fr-landmark | Facial landmark detection |
• 68-point facial landmark detection • Facial geometry analysis • Feature point extraction |
| fr-analyze | Face analysis (blur detection, orientation, etc.) |
• Image quality assessment • Head pose estimation • Eye state analysis (open/closed) |
| fr-live | Real-time face recognition from camera feed |
• Camera integration • Low-latency processing • Live face tracking |
| fr-video | Video processing for face recognition |
• Temporal face tracking • Timeline analysis • Appearance/disappearance events |
| fr-system | Core system functionality (coming soon) |
• Configuration management • Model versioning • Performance monitoring |
| Feature | FR-Framework | face_recognition | OpenCV | dlib |
|---|---|---|---|---|
| Face Detection | ✓ (HOG & CNN) | ✓ (HOG & CNN) | ✓ (Haar & DNN) | ✓ (HOG) |
| Facial Landmarks | ✓ (68 points) | ✓ (68 points) | ✓ (limited) | ✓ (68 points) |
| Face Recognition | ✓ | ✓ | ✓ (limited) | ✓ (limited) |
| Face Analysis | ✓ (extensive) | ✗ | ✓ (limited) | ✗ |
| Real-time Support | ✓ | ✗ | ✓ | ✓ (limited) |
| Video Processing | ✓ | ✗ | ✓ | ✗ |
| REST API | ✓ | ✗ | ✗ | ✗ |
| Docker Support | ✓ | ✗ | ✗ | ✗ |
| Modular Design | ✓ | ✗ | ✗ | ✗ |
| Python Support | ✓ | ✓ | ✓ | ✓ |
- Security Systems: Face-based authentication and access control
- Attendance Systems: Automatic tracking in educational or workplace settings
- Photo Cataloging: Organizing media libraries by detected individuals
- Video Analytics: Analyzing customer demographics and engagement
- Health Monitoring: Detecting fatigue or attention levels through eye tracking
- User Experience: Adaptive interfaces based on user attention and engagement
# Clone the repository
git clone https://github.com/muhkartal/FR-Framework.git
cd FR-Framework
# Install dependencies
pip install -r requirements.txt
# Install the package in development mode
pip install -e .# Build and start the services
docker-compose up -d
# Or just the API service
docker-compose up -d fr-apiFor more details on Docker deployment, see DOCKER.md.
import cv2
from fr_framework import preprocess_image, detect_faces, detect_landmarks, analyze_face
# 1. Load and preprocess the image
image_path = "path/to/group_photo.jpg"
face_image = preprocess_image(image_path)
# 2. Detect faces
detections = detect_faces(face_image, model_type="hog")
print(f"Found {len(detections)} faces")
# 3. Detect landmarks for each face
landmarks = detect_landmarks(face_image, detections)
# 4. Analyze each face
results = []
for i, (detection, landmark) in enumerate(zip(detections, landmarks)):
# Analyze the face
analysis = analyze_face(face_image, detection, landmark)
# Store results
results.append({
"face_id": i,
"position": detection.bbox,
"blur_score": analysis["blur_score"],
"brightness": analysis["brightness"],
"head_pose": analysis["head_pose"],
"eyes_closed": analysis.get("eyes_closed", None)
})
# 5. Visualize results
image = cv2.imread(image_path)
for i, (detection, result) in enumerate(zip(detections, results)):
top, right, bottom, left = detection.bbox
# Draw bounding box
cv2.rectangle(image, (left, top), (right, bottom), (0, 255, 0), 2)
# Add face ID
cv2.putText(image, f"Face #{i}", (left, top - 10),
cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
# Add analysis info
blur_text = f"Blur: {result['blur_score']:.1f}"
pose_text = f"Yaw: {result['head_pose']['yaw']:.1f}"
cv2.putText(image, blur_text, (left, bottom + 15),
cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 0), 1)
cv2.putText(image, pose_text, (left, bottom + 35),
cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 0), 1)
# Save or display the result
cv2.imwrite("output.jpg", image)
cv2.imshow("Face Analysis", image)
cv2.waitKey(0)
cv2.destroyAllWindows()# Start the Docker container
docker-compose up -d fr-api
# Use curl to detect faces in an image
curl -X POST "http://localhost:8000/detect" \
-H "accept: application/json" \
-H "Content-Type: multipart/form-data" \
-F "file=@path/to/image.jpg" \
-F "model_type=hog" \
-F "return_image=true"# Using the Python client library
from fr_framework.api.client import FRApiClient
# Create an API client
client = FRApiClient("http://localhost:8000")
# Detect faces in an image
result = client.detect_faces("path/to/image.jpg", return_image=True)
print(f"Detected {result['num_faces']} faces")
# Get the face locations
for face in result["faces"]:
print(f"Face at: {face['bbox']}")
print(f"Confidence: {face['confidence']}")For more examples, see the Tutorials and Examples sections.
The FR-Framework provides a comprehensive REST API for integration with other applications:
| Endpoint | Method | Description |
|---|---|---|
/detect |
POST |
Detect faces in an image |
/landmarks |
POST |
Detect facial landmarks |
/analyze |
POST |
Analyze facial attributes |
/video |
POST |
Process a video file |
For detailed API documentation, see the API Reference.
- Python 3.7+
- OpenCV 4.5+
- dlib 19.22+
- face_recognition 1.3+
- NumPy 1.19+
- SciPy 1.6+
- FastAPI 0.68+
- Uvicorn 0.15+
- Python-multipart 0.0.5+
- Pydantic 1.8+
- Docker 20.10+
- Docker Compose 2.0+
Contributions are welcome! Please feel free to submit a Pull Request.
- Fork the repository
- Create your feature branch (
git checkout -b feature/amazing-feature) - Commit your changes (
git commit -m 'Add some amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
# Clone the repository
git clone https://github.com/muhkartal/FR-Framework.git
cd FR-Framework
# Create a virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies
pip install -r requirements.txt
pip install -e .
# Install development dependencies
pip install pytest black flake8
# Run tests
pytest
# Format code
black .This project is licensed under the MIT License - see the LICENSE file for details.
- face_recognition library for core recognition algorithms
- dlib for face detection and landmark detection
- OpenCV for image processing functionality
- FastAPI for the API implementation
- Project Link: https://github.com/muhkartal/FR-Framework
- Developer Website: https://kartal.dev/
If you find FR-Framework helpful, please consider giving it a star ⭐