I used the FER 2013 data which included 7 classes of facial expressions: -Angry, Disgust, Fear, Happy, Neutral, Sad, and Surprise -There are about 29000 training images and around 7000 validation images
How everything works together is as follows: -Training model was developed in Jupyter Notebook -Python script was created to output the model -Python script "camera.py" gathers the live camera feed and converts to frames -Python script "main.py" pulls everything together to run using Flask app
To run the model, you need to just run python main.py in cmd prompt and go to the URL.