Face Recognition API

This project provides an API for image-based registration and recognition using the InsightFace library. It allows users to register images with associated names and recognize images against a database of registered images.

Project Structure

face-recog-api
├── src
│   ├── api
│   │   ├── main.py          # Entry point for the FastAPI application
│   │   ├── routers
│   │   │   ├── register.py  # API endpoint for registering images
│   │   │   └── recognize.py # API endpoint for recognizing images
│   │   └── schemas
│   │       └── index.py     # Pydantic models for request and response schemas
│   ├── face_recog
│   │   ├── __init__.py      # Package initialization
│   │   ├── register.py      # Logic for registering images
│   │   ├── recognize.py     # Logic for recognizing images
│   │   └── cli.py           # Command-line interface for image operations
├── tests
│   └── test_api.py          # Unit tests for API endpoints
├── Dockerfile               # Instructions to build the Docker image
├── docker-compose.yml       # Service orchestration for the application
├── .dockerignore            # Files to ignore when building the Docker image
├── requirements.txt         # Python dependencies for the project
└── README.md                # Project documentation

Setup Instructions

1. Clone the repository:

   git clone <repository-url>
   cd face-recog-api
   

2. Build the Docker image:

   docker build -t face-recog-api .
   

3. Run the application using Docker Compose:

   docker compose up
   

4. Access the API: The API will be available at http://localhost:8000. You can use tools like Postman or curl to interact with the endpoints.

API Endpoints

Register an Image

• Endpoint: POST /register
• Request Body:

  {
    "image": "base64_encoded_image_string",
    "name": "person_name"
  }

• Response:
  • Success: {"message": "Registered successfully", "id": "unique_id"}
  • Error: {"error": "Error message"}

Recognize an Image

• Endpoint: POST /recognize
• Request Body:

  {
    "image": "base64_encoded_image_string",
    "topk": 1,
    "threshold": 0.5
  }

• Response:
  • Success: {"matches": [{"name": "person_name", "id": "unique_id", "score": score}]} or {"matches": []}
  • Error: {"error": "Error message"}

Testing

To run the tests, use the following command:

pytest tests/test_api.py

License

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