This FastAPI application utilizes advanced face recognition techniques combined with a robust back-end system to manage user data and images. It supports a variety of operations, including user creation, image uploading, and retrieval of user details, making it versatile for applications requiring biometric functionality. DeepFace is used for running the Face models and encoding Face Images (a GPU is required)
Ensure you have Python 3.6+ installed on your system. You also need to have pip available for installing Python packages. Moreover you need Cmake. If you use VSCode ensure you have the C++ support installed.
Install all necessary Python packages by running the following command:
pip install fastapi uvicorn sqlalchemy opencv-python opencv-python-headless numpy python-multipart cmake jinja2 deepface tf-keras {tensorflow, tensorflow-metal or tensorflow-gpu} tqdm
These packages include FastAPI for the web framework, Uvicorn for serving the application, SQLAlchemy for database operations, OpenCV for image processing, and other supporting libraries.
This application uses SQLite as the database, which is a lightweight disk-based database that does not require a separate server process. The database file (user.db) will be created in your project directory automatically upon running the application if it does not already exist.
To start the server, use the following command:
python3 backend
To reload all the embeddings, use the following argument:
python3 backend -r
Below are the provided endpoints along with how to use them effectively:
Description: Adds a new user with the provided first name, last name, and greeting.
Usage: Send a JSON object with first_name, last_name, and greeting.
Description: Uploads and stores an image to the specified user's profile.
Usage: Form-data with image_file field containing the image file.
Description: Deletes the specified user from the database along with all associated images.
Usage: Specify the user_id in the URL.
Description: Retrieves a list of all registered users.
Usage: Direct GET request to fetch data.
Description: Fetches detailed information about a specific user by user_id.
Usage: Include user_id in the URL.
Description: Removes an indexed image from a user's profile.
Usage: Provide user_id and specify the image_index to remove.
Description: Downloads all images for a specified user compressed into a ZIP file.
Usage: Provide the user_id.
Description: Updates details for a specified user.
Usage: Send a JSON object with any updated user fields.
Description: Retrieves a specific image from a user's collection by index.
Usage: Specify both user_id and image_index.
Description: Retrieves the time the two users have spend in frame together.
Usage: Specify user1, user2 and retrieve a float in seconds.
Description: Retrieves a specific user's time he has spend in frame.
Usage: Specify user_id and retrieve a float in seconds.
Description: compare the first image of each user returning 0-1 how similar they are
Usage: Provide the user_id_1 and user_id_2
Description: Export the Table with ID, FISTNAME, LASTNAME, IMAGEID, IMAGEDATA as a python friendly csv
Usage: Connect to this endpoint from a client.
Description: Retrieves all users that are currently in frame.
Usage: Connect to this endpoint from a client.
Description: Streams a live video feed that is processed to detect and recognize faces.
Usage: Connect to this endpoint from a client that supports video streaming.
Description: Serves the homepage of the application.
Usage: Access this endpoint from any web browser to view the homepage.
This README provides all the necessary details to get the application up and running, and to understand how to interact with its comprehensive suite of endpoints. For production environments, consider additional security, performance optimization, and error handling mechanisms.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent