The trained model is contained in mask-detect/trained_model/saved model/. Do not modify any of the files in this directory. These are the files used to define our ML model.
To run the code, you need tensorflow and numpy installed.
Tensorflow installation instructions can be found here: https://www.tensorflow.org/install/pip
Note that Tensorflow recommends the use of Python virtual environments. It's not mandatory, but it will help to manage potential conflicts between dependancies.
The command below willl installl numpy. For more installation instructions: https://numpy.org/install/
pip install numpy
Ensure that mask-detect/trained_model/test_files/ contains either PNG or JPG images. The ML model will test the images in this directory. You can find some sample files to use in the shared Capstone Google Drive: capstone/images/test_run_images
To run the trained model, run the mask-detect/trained_model/model_run.py file. You may need to use either python mask-detect/trained_model/model_run.py or python3 mask-detect/trained_model/model_run.py, depending on the python version
When run,model_run.py will look at each of the fles in mask-detect/trained_model/content/, and output whether each image is of someone with or without a face mask.
This repo is using GitHub LFS (Large File Storage), which only provides 1 GB of data transfer a month. To avoid going over the data cap, please do not upload any images to the respoitory.
train.zip and valid.zip contain the training and validation data used to train the ML model. The code used to train the ML model depends on these files, so please do not modify them.