Summary

International Symposium on Nonlinear Theory and its Applications

2008

Session Number:A2L-D

Session:

Number:A2L-D3

Sequential Superparamagnetic Clustering as a Predictor of Visual Fixations

Tom Jasa,  Thomas Lanz ,  Thomas Ott,  Ruedi Stoop,  

pp.-

Publication Date:2008/9/7

Online ISSN:2188-5079

DOI:10.34385/proc.42.A2L-D3

PDF download (145.6KB)

Summary:
Clustering of different visual patterns from background noise is a task which is easily performed by the human visual system but still proves to be difficult problem for artificial systems. The Sequential Superparamagnetic Clustering (SSC) algorithm can be used to cluster the visual patterns and provides a relative measure of stability for each cluster. By using an eye tracking aparatus which measures the human visual scan paths, a comparisson between statistical properties of the scan paths, and the relative stability for the clusters given by the SSC algorithm is determined. This comparisson of visual scan paths and relative stability provides insight into how the human visual system performs the task of clustering and quantifies how similar the SSC algorithm is to the human visual system.