The project classifies the facial emotions in real time. The model file is trained on the following emotions: “angry”, “disgust”, “fear”, “happy”, “sad”, “surprise”, “neutral”. A 21 layered architecture was trained on FER 2013 dataset and it obtained an accuracy of 68%. For face detection HAar Cscade classifier was implemented.
Dependencies: Keras, Numpy, Pandas, OpenCV, Tensorflow, Statistics.
To train run Train.py and for testing run Test.py.
Model files for Face Classifier and Emotion Classifier are also provided.
Reference Paper: Arriaga, Octavio, Matias Valdenegro-Toro, and Paul Plöger. "Real-time Convolutional Neural Networks for Emotion and Gender Classification." arXiv preprint arXiv:1710.07557 (2017).
Link for Dataset: https://www.kaggle.com/c/challenges-in-representation-learning-facial-expression-recognition-challenge/data