Facial Expression Detection

A project on creating a system to detect the facial emotion expressed by a human. This system can run on video streamable platforms. A new CNN was trained without using transfer learning for building of model. In this project we can see the effects of 2 different optimizers during the training of the model. This repository also contains a code for the UI to implement the model.

System Architecture:

Model Summary:

Output:

From YouTube Videos:

For Picutres:

Note:

The haarcascade_frontalface_default.xml file is required to run the modle as that file is used by the model to identify the face.

Authors

Vishnu S S, Aniruddh T S