Automated Recognition of Facial Expressions
AUTHORS:H.J. Huisman
KEYWORDS:
ABSTRACT:
This is the graduation report, written to gain a master degree in Computer
Science at the Delft University of Technology. It covers eight months of
theoretical and practical work: researching vision literature, experimenting
with the backpropagation learning paradigm, implementing and testing the
Neocognitron (a complex neural network for image processing) and recording
and analysing results. The goal is to find a neural network capable of
detecting emotions from pictures. This graduation project is only a small
part of larger research project, aiming at automated detection of emotions
and particularly stress with several media: vision, sound, and physiology.
Only the vision medium is subject of this study.
