Navigating with and without Landmarks in Virtual Reality

Vlada Aginsky

Department of Cognitive and Linguistic Science,
Brown University

Abstract:
Humans and animals are able to switch from one navigational strategy to another, depending on what type of environmental information appears most reliable in a given situation. For instance, hamsters will rely on visual landmarks until the landmarks are displaced so far from their original location that they offer exceedingly divergent information from path integration. At that point they switch to the path integration strategy, which entails using proprioceptive signals during locomotion to continuously update the animal’s position relative to some starting point. In this study our aim was twofold. First of all we were interested in whether or not people will use path integration when landmarks are available but are known to be less reliable than internally-generated signals. Secondly, if people do decide to rely on visual landmarks – will they choose to focus on a particular type of landmark, local or global? Using a virtual reality environment, subjects moved through a tunnel containing a 115 degree bend onto an arena surrounded by equally-spaced white poles, one of which was the target pole. The task was to go to the target pole (colored red during training, white during test) which remained the same throughout the entire experiment. The environment also contained three rectangular objects inside the arena (local landmarks) and three larger ones outside the arena (global landmarks). For each trial one local, one global, all local, all global, or all objects shifted 6, 12, or 18 degrees to the right or left of their original location. The subjects were warned of this. Results showed that despite their knowledge that objects could move, subjects relied on them as landmarks to find the target pole. Furthermore they relied more on local objects than global objects. These results will be contrasted with subjects’ performance in an environment without landmarks.

The lecture will take place:

in the Lecture Hall, Room 203, 44 Cummington St.
on Monday, March 27, 2000
at 1:00 pm