4th Annual Meeting of the International Multisensory Research Forum
    Home > Papers > Florian Roehrbein
Florian Roehrbein

Auditory-visual motion signals and the statistical redundancy of the natural environment
Poster

Florian Roehrbein
Institute for Medical Psychology, Ludwig-Maximilians-Universitaet Muenchen, Germany

Christoph Zetzsche
Institute for Medical Psychology, Ludwig-Maximilians-Universitaet Muenchen, Germany

Georg Meyer
Department of Psychology, University of Liverpool

     Abstract ID Number: 108
     Full text: Not available
     Last modified: May 20, 2003

Abstract
The sensory stimulation that results from the typical observer behaviour in a natural environment causes systematic co-variations, especially relative movements can cause redundancies between the auditory and visual channel. Since an independent, modality-specific processing would be sub-optimal, we ask here whether the perceptual system is capable of exploiting such intermodal redundancies by an optimized joint processing strategy.
We tested this idea in two experiments. First we measured auditory-visual incremental thresholds for a compound stimulus consisting of a lightspot and a 1kHz tone. We observed an increased sensitivity for ecological relevant stimulus configurations with joint increments or decrements for both tone and light. In a second experiment moving audio-visual stimuli describing an arc of 90 degrees were presented in one hemi-field. The auditory and visual components could move independently, hence they could be in the same or different hemi-fields and move in the same or opposite direction. Again, an increased sensitivity was observed only for the ecologically relevant configurations in which the motion signals are co-localized.
We conclude that there exists a selective neural integration mechanism for ecologically relevant audio-visual stimulus combinations. Our results thus provide strong behavioural evidence against separate information processing in early sensory channels and they suggest that the adaptation to environmental statistical regularities is a general principle of neural information processing.


    Learn more
    about this
    publishing
    project...


Public Knowledge

 
Open Access Research
home | overview | program
papers | organization | links
  Top