Code for Build Your Own Face Recognition Tool With Python

This is a small project to demonstrate how to build a face recognition tool using Python. It goes along with the Real Python tutorial Build Your Own Face Recognition Tool With Python.

Dependency Installation

dlib requires both CMake and a C++ compiler to be installed before using it. Refer to the CMake documentation and that of your preferred C++ compiler for installation instructions.

The script assumes that the directories training/, output/ and validation/ exist, so be sure to create those directories before running the code.

Usage

The three major phases of the machine learning workflow are represented by the program's three switches:

--train: Initiate the model training process.
--validate: Run a trained model on images stored in the validation directory. The faces in these images should be known to you so you can debug the model's performance.
--test: Run a trained model on an image with an unknown face in it. The script will use the model to detect and attempt to identify the face in the image.
-m: Specify the type of model architecture you want to use. "hog" is the default and best for CPU-based training, while "cnn" is better for GPU training and will generally give better performance.