• DocumentCode
    3325282
  • Title

    Neural preprocessing of auditory-wind sensory signals and modular neural control for auditory- and wind-evoked escape responses of walking machines

  • Author

    Manoonpong, Poramate ; Wörgötter, Florentin ; Pasemann, Frank

  • Author_Institution
    Bernstein Center for Computational Neuroscience, University of Göttingen, Bunsenstr. 10, 37073, Germany
  • fYear
    2009
  • fDate
    22-25 Feb. 2009
  • Firstpage
    786
  • Lastpage
    793
  • Abstract
    The flying crickets Teleogryllus oceanicus have sound sensitive organs to elicit an “acoustic startle response”. Another kind of the startle response is also evident in the cockroaches Periplaneta and the crickets Gryllus bimaculatus where they use their cercal filiform hairs (wind sensitive hairs) to trigger so called “wind-evoked escape behavior”. Similarly, in this paper a setup is described where an auditory-wind detector sensor and a neural preprocessing system together with a modular neural controller are used to simulate such animal behaviors in an abstract form on a six-legged walking machine. The neural preprocessing network acts as a low-pass filter and sensory shaping unit. In addition, the modular neural controller then generates the desired responses such that the machine performs a fast movement away from abrupt, intense, and unexpected auditory and/or wind stimulus in a physical environment.
  • Keywords
    Acoustic sensors; Animal behavior; Detectors; Hair; Legged locomotion; Personal digital assistants; Robot sensing systems; Sense organs; Sensor systems; Ultrasonic imaging; Autonomous hexapod robots; Biologically-inspired robots; Neural control; Recurrent neural networks;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Robotics and Biomimetics, 2008. ROBIO 2008. IEEE International Conference on
  • Conference_Location
    Bangkok
  • Print_ISBN
    978-1-4244-2678-2
  • Electronic_ISBN
    978-1-4244-2679-9
  • Type

    conf

  • DOI
    10.1109/ROBIO.2009.4913100
  • Filename
    4913100