The Performance Of Automated Facial Expression Analysis
Keywords
Loading...
Authors
Issue Date
2017-08-29
Language
en
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This research is about the performance of automated facial expression recognition.
One of the goals was to examine whether an open source system
can compete with a state of the art (and expensive) system. In order to do
this a combination of open source systems DRMF2013 and WEKA was compared
to FaceReader, by running them on several images of different data
sets. The used pictures were divided in two categories; the first with frontal
images only and the second with images where persons had their heads in positions
between 45 degrees to the left and the same degrees to the other side.
The results show that the developed open source system can compete with
FaceReader, as it beats FaceReader when it comes to the group of frontal images
and the images with a greater angel. However, when it comes to frontal
images only, FaceReader beats the open source system. These findings are
significant, as money could be saved in research projects that for example
involve automated facial expression recognition.
Description
Citation
Supervisor
Faculty
Faculteit der Sociale Wetenschappen