A Comparison of Attention Estimation Techniques in a Public Display Scenario
Sprache des Vortragstitels:
19th International Conference on Human-Computer Interaction (HCI), Vancouver, Canada, July 9-14, 2017.
Sprache des Tagungstitel:
Human interaction with a public display presupposes a person?s attention. An Interactive display, hence, aim sat attracting attention by e.g. emitting a strong signal that makes the inattentive visitor turn towards it. The challenge in this regard is to reliably determine the attention of passers-by. In this article, we investigate different technical methods for estimating attention in a public display scenario by measuring physical expressive features, from attention can be derived. In the course of an experimental setup we compare a Support Vector Machine, a neural network using a Multilayer Perceptron and a Finite State Machine and compare the results to a manual reference classification. We carve out strengths and weaknesses and identify the most feasible measuring method with regard to precision of recognition and practical application.