Navigation

Christopher Fichtel

  • Job title: Bachelor's Thesis
  • Working group: Stress Classification based on Movement Patterns in the Stroop Room

Advisors
Stefan Gradl, M. Sc., Markus Wirth, M. Sc., Prof. Dr. Björn Eskofier

Duration
12/2020 – 05/2020

Abstract

The Stroop color word test has already been used in studies to detect and analyze stress and
anxiety. For the classic test, which includes speech of the participants, it could be shown that facial
cues, such as head movements, can provide insight into the analysis and classification of stress.
However, the exact correlation between head movement and stress has not been established yet [1].
The head displacement may also be caused by other body functions, e.g. speech [2]. Nevertheless it
is well known that stress induces more frequent, more rapid and overall more head movement [1, 3].
Using a virtual reality implementation of the Stroop color word test, the Stroop Room [4],
movements of the head as well as of the hand holding the controller will be analysed and evaluated
regarding their potential as stress indicators. First, periods of stress will be manually classified and
labeled using the data of recorded biosignals. Based on the position data of head and controller,
features will be extracted for a stress classifier. These should show differences for the congruent
and incongruent phase. It might even be possible to extract micromovements using adequate
filtering or more sophisticated post-processing techniques [2]. The goal of the thesis is to develop
a machine learning pipeline using movement patterns as a stand-alone stress indicator [5].

 

References:

  1. Giannakakis, G. et al.:Stress and anxiety detection using facial cues from videos. Biomedical
    Signal Processing and Control , 2017.
  2. Giannakakis, G. et al.: Evaluation of head pose features for stress detection and classification. IEEE EMBS International Conference on Biomedical & Health Informatics (BHI),
    2018.
  3. Sharma, N and Gedeon, T..: Objective measures, sensors and computational techniques
    for stress recognition and classification: A survey. Computer Methods and Programs in
    Biomedicine 108 (1287-1301), 2012.
  4. Gradl, Stefan, Markus Wirth, Nico Machtlinger, Romina Poguntke, Andrea Wonner, Nicolas
    Rohleder, and Bjoern M. Eskofier. 2019. “The Stroop Room: A Virtual Reality-Enhanced
    Stroop Test.” In 25th ACM Symposium on Virtual Reality Software and Technology –
    VRST ’19, 1–12. Parramatta, NSW, Australia: ACM Press. https://doi.org/10.1145/
    3359996.3364247.
  5. Duda, Richard O, Peter E Hart, and David G Stork. 2012. Pattern Classification. New York,
    NY: John Wiley & Sons. http://nbn-resolving.de/urn:nbn:de:101:1-2014120312432.
  6. AutoML in Python. Accessed 2019-12-13. https://heartbeat.fritz.ai/
    automated-machine-learning-in-python-5d7ddcf6bb9e