Visuell versus bedeutungsvoll - the Miracle of Art
Elisa Tersteegen
University of Vienna · 2010
The aim of this study was to explore whether during the perception of art works visual cues or the content of a picture were better predictors of the viewers´ eye movements. Visual cues were defined according to the Saliency Model by Itti and Koch (2000). To define the most important areas of 30 modern, narrative paintings and to generate a resultant importance map, a prestudy was conducted.
In order to compare the quality of the fixation predictions made by the saliency maps and the importance maps, an eye-tracking study was conducted. The 30 paintings were presented to 17 participants. The results indicate that the content of the paintings is a better predictor for fixations than the saliency maps, with respect to average fixation duration, number of fixations, total fixation duration and first AOI fixation.
Hence, during the perception of artworks, the content of a picture seems to guide the viewers´ eye movements from the first glance.