Human emotion classifier - Personal Project

Description

This project is a ML-based emotion classifier meant to demonstrate my proficiency with multiple Python machine learning libraries as well as my ability to construct and fit a neural network to data. This high-level idea is that this program connects to a device that feeds in a stream of input, such as a laptop camera, and takes that input and uses the model to predict what emotion is being expressed by the stream. The Model training and construction are built using Python3 tensorflows Keras and the camera stream is preprocessed for the model thanks to the c compilation of the numpy Pyton3 package.

To run

cd NN_traing
python3 train.py

Disclaimer

The data the model was trained on was provided as open-source data thanks to Kaggle. You can find the link to the data as a comment in the top of the file "train.py".