Python version: 3.7
The following python packages are required
- numpy
- openCV
- pandas
- matplotlib
- tensorflow/keras
Requirements: To run this script, the fer2013 data set must be downloaded (e.g., https://www.kaggle.com/deadskull7/fer2013) and placed in the 'src' folder. By default, this script loads the 'finalNet.h5' model and predicts images from the fer2013.csv data set.
The file 'haarcascade_frontalface_default.xml' inside the 'src' folder is required to execute this script. This script can be executed to predict an image (e.g. testImage.png) by passing the image as an argument:
python predictImage.py testImage.png