Cite as:
Kayahara T, Kitazaki M, Yoshino T, Ogi T, Hirose M, 2003, "Effect of observer's self-motion sensation on scene recognition in immersive virtual-reality environment" Perception 32 ECVP Abstract Supplement

Effect of observer's self-motion sensation on scene recognition in immersive virtual-reality environment

T Kayahara, M Kitazaki, T Yoshino, T Ogi, M Hirose

In this study, the recognition of objects' layout across views was examined in the immersive virtual-reality (VR) environment called CABIN. It has five screens (2.5 m each) which are arranged like a cube surrounding an observer. An observer views the image dichoptically through LCD shutter glasses. Thus, the observers can immerse into 3-D VR environment. Using one screen 2-D VR environment, we have reported that the performance of scene recognition depended on visually induced self-motion sensation [Kitazaki and Yoshino, 2002 Journal of Vision 2(7) 633a]. The purpose of the study was to examine whether our scene representation is renewed by the visual information of the self-motion in the immersive 3-D VR. We constructed a virtual room in CABIN and set a table at the centre of the room, on which five geometrical objects were randomly located. A trial consisted of a memorisation phase (3 s), a retention phase (7 s), and a recognition phase (3 s). In the retention phase, the table and/or the room were rotated 47°, while the objects on the table were occluded and one of them moved to a new position. Room rotation simulated viewpoint movement. The observers were required to identify the moved object. When recognition was made from same viewpoint (without room rotation), the performance was more accurate in the condition of the same retinal image of the layout of the object (without table rotation) than of a different retinal image. However, when recognition was made from a different viewpoint (with room rotation), the performance was more accurate with the different retinal image, which was consistent with viewpoint movement in the real-world. These results suggest that our scene representation can be renewed automatically and inevitably by visual self-motion simulation in the immersive 3-D VR.[This research was supported by SCAT, Nissan Science

These web-based abstracts are provided for ease of seaching and access, but certain aspects (such as as mathematics) may not appear in their optimum form. For the final published version of this abstract, please see
ECVP 2003 Abstract Supplement (complete) size: 1712 Kb