非表示:
キーワード:
-
要旨:
Exaggeration techniques are an excellent tool for investigating the representation of complex visual stimuli. In particular, they can help explore the notion that such stimuli are encoded as elements within continuous multidimensional perceptual spaces. For example, experimental evidence supporting this idea has been provided by the caricature effect for faces (eg Rhodes et al, 1987 Cognitive Psychology 19 473 - 497; Benson and Perrett, 1994 Perception 23 75 - 93). Traditionally, exaggeration techniques have been applied to static stimuli. However in everyday life we are mostly confronted with moving stimuli. Can complex-motion stimuli, such as biological movements, also be represented in continuous multidimensional spaces? Several methods have recently been developed which suggest that this is possible. These methods can be used to produce spatiotemporal caricatures. The aim of this contribution is twofold. First, we give an overview of these various spatiotemporal exaggeration methods. Second, we present a new method for the morphing and exaggeration of long complex sequences of facial expressions using photorealistic head models. In addition, preliminary data are shown which demonstrate the effect of spatiotemporal exaggeration on the ability to discriminate between complex individual facial-motion patterns.
