Prerequisites

• scikit-learn
• opencv-python
• keras
• tensorflow
• tensorflow-gpu (if a GPU is available)
• numpy
• Python 3.6.4+

Building the dataset

$ python3 make_dataset.py

The main_dir variable must be edited depending on the actual path of the dataset.

Optionally, you can augment the images prior to making the dataset.

$ python3 augment-data.py

Training the convnet

$ python3 tf-train.py

Note that this can take quite a long time without a GPU, espeically after augmenting the data. To alleviate this, the number of epochs can be decreased.

You can test the model independently by running

$ python3 test-data.py

Running the webcam applet

python3 webcam.py