Towards attention-centered interfaces

Abstract

The established method of representing three-dimensional space on a two-dimensional surface involves camera based, point of regard systems, comparable in design to the early “camera obscura”. However, geometrical limitations of such models lead to distortions of perspective when projected. This research investigated the influence of single- versus multi-perspectives on aesthetic choices within one image. A clear perceptual bias towards multi-perspective images was found, additionally supported by an eye tracking study. We propose that human users are more attracted by multi-perspective images, which emphasise the “semantic foci” of the scene, than by those being synthesized statically with only one geometrical prospect.