In visual navigation, landmarks can be used in a number of different ways. In this paper, we investigate the role of global and local landmarks in virtual environment navigation. We performed an experiment in a virtual environment called ''Hexatown,'' consisting of a regular hexagonal grid of streets and junctions. Each junction was identi ed by the presence of distinct local landmarks (buildings, phone box, and so on). Additionally, compass information or a global frame of reference was provided by global landmarks (hilltop, television tower, and city skyline). According to participants' movement decisions, egomotion was simulated, and displayed on a 180 deg. projection screen. Participants learned the route back and forth between two local landmarks. In the test phase, individual junctions were approached and the participant's movement decision was recorded. We performed two experiments involving landmark changes after learning. In the rst, we used con icting cues by transposing landmarks. In the second experiment, we reduced either local or global landmark information. Results show that both local and global landmarks are used in way nding decisions. However, different participants rely on different strategies. In the rst experiment (cue con ict) for example, some of the participants used only local landmarks while others relied exclusively on global landmarks. Other participants used local landmarks at one location and global landmarks at the other. When removing one landmark type in the second experiment, the other type could be used by almost all participants, indicating that information about the neglected landmark type was present in memory.

No abstract

Theoretical considerations and earlier experimental findings indicate that traveling over slanted terrain can lead to an enrichment of the perceived spatial cues relevant for navigation. We investigated the proposed facilitation of a uniformly slanted environment on navigation and orientation performance with a virtual environment presented on a large 180 degrees screen, using as material a virtual town with eight places and twenty-four landmarks. In the control condition, this town was placed on a flat surface; in the two experimental conditions, the town was placed on a slope with a uniform angle of 4 degrees. Pedaling on a bicycle simulator, participants first explored the environment, then solved navigation tasks, pointed from various positions to distant landmarks, judged the relative elevation of pairs of distant landmarks from memory, and finally drew a sketch map of the environment. In comparison to the control condition, the number of navigation errors was significantly lower in the slanted conditions, and the deviations in the pointings to distant landmarks were massively reduced. Participants from the slant conditions also showed good knowledge of the relative elevations of pairs of distant locations. However, no differences in map-drawing quality were found. The results lend additional support to the proposition that our spatial knowledge, which is used in navigation and orientation, contains vertical information.

We investigatedthe role of geographical slant in simple navigation andspatial memory tasks, using an outdoor virtual environment. The whole environment couldb e slantedb y an angle of 4◦. Subjects could interact with the virtual environment by pedaling with force-feedback on a bicycle simulator (translation) or by hitting buttons (discrete rotations in 60◦ steps). After memory acquisition, spatial knowledge was accessed by three tasks: (i) pointing from various positions to the learnedgoals; (ii) choosing the more elevatedof two landmarks from memory; (iii) drawing a sketch map of the environment. The number of navigation errors (wrong motion decisions with respect to the goal) was significantly reducedin the slantedcond itions. Furthermore, we foundthat subjects were able to point to currently invisible targets in virtual environments. Adding a geographical slant improves this performance. We conclude that geographical slant plays a role either in the construction of a spatial memory, or in its readout, or in both

No abstract