Traditionally, perceptual research has been compartmentalized into distinct and isolated categories according to individual modalities (i.e., visual, auditory, haptic, proprioceptive, vestibular, etc.).  This modular approach has treated individual sensory and motor processing as involving largely independent systems.  However, recently investigators are recognizing the importance of understanding cross-modal interactions and how they relate to perception and behaviour.

Much of the multisensory research up until this point has focused upon tasks involving discrete stimulus presentations in near body space. Such cue interactions of interest have included: visual-auditory integration, visual-proprioceptive integration, or visual-haptic integration. It is important, however, to investigate multisensory integration from the perspective of large-scale self-motion through action space. Unlike traditional approaches to examining the integration of two specific cues at a particular instance in time, navigating through one’s environment requires the dynamic integration of several cues across space and over time (i.e., visual flow, lower-limb proprioception, and vestibular information).  Understanding the principles underlying multimodal integration in this context of unfolding cue dynamics is very important as it provides insight into an important category of multisensory processing.

◘ Multi-sensory integration in the estimation of distance traveled
◘ Contribution of inertial information to the perception of walking speed
◘ Perception of visual speed while walking
◘ Adaptive treadmill control
◘ Vestibular perception is slow
◘ Perceived object stability
◘ Shape from shading

Traversed distance perception involves estimating the extent of self-motion as one travels from one position in space to another. As such, it is a multi-modal experience in which information from both visual flow and non-visual cues (i.e. proprioceptive, efference copy and vestibular cues) jointly specify the magnitude of self-motion. While recent evidence has demonstrated the extent to which each of these cues can be used independently to estimate traversed distance, relatively little is known about how they are integrated when simultaneously present.

This experiment required participants to move down an infinitely long virtual hallway and subsequently judge how far they walked.  Participants either experienced the movement using vision alone, non-visual cues alone (walking without vision), or both cues combined.  In the combined condition, the visual and non-visual cues were either congruent or in conflict. 

Results demonstrated that both visual and non-visual cues contribute to one’s estimate of traveled distances.  Specifically, the combined cue condition demonstrated results that differed than those observed for either cue alone.  However, when the two cues were placed in conflict, locomotor cues appear to be weighted higher overall.

REFERENCE
◘ Campos, J., J. S. Butler, B. Mohler and H. H. Bülthoff: The contributions of visual flow and locomotor cues to walked distance estimation in a virtual environment. APGV; July, 2007 (link to document)

The virtual environment

Subjects wear a tracked HMD and
a backpack