Implementation in Tensorflow 2.x of Engagement Recognition using DAiSEE dataset
Engagement Recognition is made through a multi-label classification of the following status:
- engaged
- bored
- frustrated
- confused
This implementation comes with a camera script so to test it using your webcam.
Download the dataset running the .sh script
You can install dependencies as usual using pip install -r requirements.txt Then you need to preprocess the dataset by running daisee_data_preprocessing.py script. Afterwards run train.py to train the model and test it using main.py script