In two experiments, subjects traveled through virtual mazes, encountering target objects along the way. Their task was to indicate the direction to these target objects from a terminal location in the maze (from which the objects could no longer be seen). Subjects controlled their motion through the mazes using three locomotion modes. In the Walk mode, subjects walked normally in the experimental room. For each subject, body position and heading were tracked, and the tracking information was used to continuously update the visual imagery presented to the subjects on a head-mounted display. This process created the impression of immersion in the experimental maze. In the Visual Turn mode subjects moved through the environment using a joystick to control their turning. The only sensory information subjects received about rotation and translation was that provided by the computer-generated imagery. The Real Turn mode was midway between the other two modes, in that subjects physically turned in place to steer while translating in the virtual maze; thus translation through the maze was signaled only by the computer-generated imagery,

this article. Because there is no visible cue in the hidden-platform water maze task, it would not help the animal find the platform. 3. Route system. Routes stored in the hippocampus can be written out to the cortex, so that directions necessary to reach a goal are associated with local views. This is the system detailed in section 2.5 (see also section 4.3). This system requires training for each step the animal must take; it cannot learn to associate local views with directions to distant goals without hippocampal help (through route replay). The Role of the Hippocampus 97 If there were a way to show the animal the route to the goal, it might be possible to train the route system even without a hippocampus. Whishaw, Cassell, and Jarrard (1995) and Schallert, Day, Weisend, and Sutherland (1996) both showed ways to train the route system directly and found that animals could learn to solve the water maze even with hippocampal lesions. Whishaw et al. (1995) trained animals with fimbria/fornix lesions to find a visible platform and then removed the visible platform. These animals concentrated their search where the platform had been. Schallert et al. (1996) used animals with kainate-colchicine hippocampal lesions. The animals were first trained with a large platform that filled almost the entire maze. Once the animals could reach that platform reliably, it was shrunk trial by trial until it was the same size as a typical platform in a water maze task. Again, the animals could learn to solve the water maze without a hippocampus. 4.3 Where Is the Route System? Although the data are not yet conclusive, we suggest that the most likely candidate for anatomical instantiation of the route system is from posterior parietal to posterior cingulate cortex. There is a lot of evide...

A computer model of rodent navigation, based on coupled mechanisms for place recognition, path integration, and maintenance of head direction, offers a way to operationally combine constraints from neurophysiology and behavioral observation. We describe how one such model reproduces a variety of experiments by Collett, Cartwright, and Smith [6] in which gerbils learn to find a hidden food reward, guided by an array of visual landmarks in an open arena. We also describe some neurophysiological predictions of the model; these may soon be verified experimentally. Portions of the model have been implemented on a mobile robot. 1. Introduction Landmark-based navigation is a rich domain for exploring issues of visual and spatial cognition. At the behavioral level, there is a wealth of data on how animals use landmarks to locate food or return to their nests. At the neurophysiological level, hippocampal pyramidal cells called place cells have been discovered that fire when the animal is in a ...

this article should be addressed to either Sergio S. Fukusima, Department of Psychology, FFCLRP, University of Sat Paulo, Ribeiro Preto, Sat Paulo, Brazil, CEP 14050-901, or Jack M. Loomis, Department of Psychology, University of California, Santa Barbara, California 93106-9660. Electronic mail may be sent via Intemet to fukusima@usp.br or loomis @ psych.ucsb.edu. distance is not linked tightly to that of exocentric distance (see also Gogel, 1977), our focus here is on the former. It generally is accepted that when visual cues to distance are reduced greatly, egocentric distance is misperceived (e.g., Baird, 1970; Da Silva, 1985: Foley, 1977, 1980; Foley & Held, 1972; Gogel, 1974; Holway & Boring, 1941; Kiinnapas, 1968; Philbeck & Loomis, 1997; Sedgwick, 1986). Under "full-cue" conditions, in which a stimulus-rich envi- ronment is viewed under good illumination, however, there is little agreement about whether perception is accurate, mainly because of the diversity of findings stemming from different experimental methods. With respect to egocentric distance, much of the research conducted under full-cue conditions suggests that perceived distance is nearly linear in physical distance and appropriately scaled, at least for targets within 20 m (e.g., verbal reports, Da Silva, 1985; Sedgwick, t986; Teghtsoonian & Teghtsoonian, 1969, t970; blind walking to previewed targets, Corlett, Patla, & Williams, 1985; Elliott, 1986, 1987: Elliott, Jones, & Gray, 1990; Loomis, Da Silva, Fujita, & Fukusima, 1992; Rieser, Ashmead, Talor, & Youngquist, 1990; Steenhuis & Goodale, 1988; Thomson, 1983); in those studies, power functions with exponents close to 1.0 were obtained. The results of other research under the same viewing conditions and over the' same physical distances suggest a c...

The ability to acquire a representation of the spatial environment and the ability to localize within it are essential for successful navigation in a-priori unknown environments. The hippocampal formation is believed to play a key role in spatial learning and navigation in animals. This paper briefly reviews the relevant neurobiological and cognitive data and their relation to computational models of spatial learning and localization used in mobile robots. It also describes a hippocampal model of spatial learning and navigation and analyzes it using Kalman filter based tools for information fusion from multiple uncertain sources. The resulting model allows a robot to learn a place-based, metric representation of space in a-priori unknown environments and to localize itself in a stochastically optimal manner. The paper also describes an algorithmic implementation of the model and results of several experiments that demonstrate its capabilities.

Abstract not found

Although evolutionary algorithms have been employed to automatically synthesize control and behavior programs for robots and even design the physical structures of the robots, it is impossible for evolution to anticipate the detailed structure of specific environments that the robot might have to deal with. Robots must thus possess mechanisms to learn and adapt to the environments they encounter. One such mechanism that is of importance to mobile robots is that of spatial learning, i.e., the ability to learn the spatial locations of objects and places in the environment, which would allow them to successfully explore and navigate in a-priori unknown environments. This paper proposes a computational model for the acquisition and use of spatial information that is inspired by the role of the hippocampal formation in animal spatial learning and navigation. 1 Introduction Mobile robotics has progressed significantly in the last four decades leading to numerous applications in mail-deliv...

In two experiments, subjects traveled through virtual mazes, encountering target objects along the way. Their task was to indicate the direction to these target objects from a terminal location in the maze (from which the objects could no longer be seen). Subjects controlled their motion through the mazes using three locomotion modes. In the Walk mode, subjects walked normally in the experimental room. For each subject, body position and heading were tracked, and the tracking information was used to continuously update the visual imagery presented to the subjects on a head-mounted display. This process created the impression of immersion in the experimental maze. In the Visual Turn mode subjects moved through the environment using a joystick to control their turning. The only sensory information subjects received about rotation and translation was that provided by the computer-generated imagery. The Real Turn mode was midway between the other two modes, in that subjects physically turned in place to steer while translating in the virtual maze; thus translation through the maze was signaled only by the computer-generated imagery, whereas rotations were signaled by the imagery as well as by proprioceptive and vestibular information. The dependent measure in the experiment was the absolute error of the subject's directional estimate to each target from the terminal location. Performance in the Walk mode was significantly better than in the Visual Turn mode but other trends were not significant. A secondary finding was that the degree of motion sickness depended upon locomotion mode, with the lowest incidence occurring in the Walk mode. Both findings suggest the advisability of having subjects explore virtual environments using real rotations and translations in tasks invol...