Projekt

Mammals have an extraordinary ability to identify auditory objects, like communication calls and speech. Echolocating mammals identify real, three-dimensional objects exclusively through the auditory ana- lysis of echoes. Mammals reliably recognize auditory objects despite the variability of the sounds from one occurence to the next. Much of the information about object shape and structure is represen- ted in its impulse response, and much of the variability relates to the size of the sound source, e.g. a child vs. an adult, or, in echo- location, a small prey vs. a large prey of the same shape. For reli- able object recognition, the auditory system must segregate informa- tion about object size from information about object structure. This results in a size-invariant object representation; the auditory object is normalised. As yet there is no proper understanding of the extra- ordinary ability of the mammalian auditory system which performs this procedure virtually automatically. This joint project focuses ex- pertise from the full breadth of auditory neuroscience on auditory ob- ject normalisation. The experimental techniques range from psychophy- sics to brain imaging and electrophysiology; the results will supply the basis for a physiologically plausible computer model of auditory object recognition.

• Prof. Dr. Lutz Wiegrebe

  Universität München
  Dept. Biologie
  Division of Neurobiology
  Martinsried
  

• Prof. Dr. Benedikt Grothe

  Universität München
  Department Biologie II
  Planegg-Martinsried
  

• Prof. Dr. Gerd Schuller

  Universität München
  Department Biologie II
  München
  

• Dr. Roy Patterson

  Medical Research Council
  Centre for the Neural Basis of Hearing
  Physiology Department
  Cambridge, CB2 3 EG
  Grossbritannien
  

• Dr. Timothy Griffiths

  University of Newcastle
  The Medical School
  Neurology, Neurobiology & Psychiatry
  Newcastle upon Tyne
  Grossbritannien