Project for Intelligent Systems - University of Pisa.
The tasks performed in this project are the following:
-
Design and develop two MLP artifcial neural networks that accurately estimate a person's valence and arousal and two RBF networks that do the same thing as the MLPs.
-
Design and develop a fuzzy inference system to estimate a person's arousal.
-
Fine-tune a pretrained CNN to obtain a convolutional neural network (CNN) that accurately classifes a person's emotion, based on facial expression.