Fly optomotor behaviour



Projects: Fly head	Fly optomotor behaviour Stabilising a Fly's Head Flies stabilise their heads very quickly when disturbed in flight (the visual reflex takes approximately 25ms). Recent research indicates that this is thanks to a very simple mapping of motion detectors onto motor neurons which innervate the fly neck muscles. This projects proposes to build a simple implementation of this system as a proof of principle and, if time allows, to investigate issues related to the mapping and connectivity of neurons in the system. We have obtained from a collaborator a silicon implementation of the elementary motion detectors in the fly, and the project would take this, together with a small motor platform and microcontroller, and build a complete robotic system using (and adding to) software we are currently developing for real-time robot control. Involved inside group: Matthew Prowse Richard Reeve Barbara Webb Involved outside group: Reid Harrison, Department of Electrical and Computer Engineering, University of Utah Recent publications: Webb, B. and Reeve, R. Reafferent or redundant: How should a robot cricket use an optomotor reflex? Adaptive Behavior (2003), 11(3):137-158 (pdf) Related papers: Estimation of self-motion by optic flow processing in single visual interneurons, Holger Krapp and R. Hengstenberg, Nature (1996) v.384, pp.463-466 A silicon implementation of the fly's optomotor control system, Reid Harrison and Christof Koch, Neural Computation (2000) v.12, pp.2291-2304

Stabilising a Fly's Head

Flies stabilise their heads very quickly when disturbed in flight (the visual reflex takes approximately 25ms). Recent research indicates that this is thanks to a very simple mapping of motion detectors onto motor neurons which innervate the fly neck muscles. This projects proposes to build a simple implementation of this system as a proof of principle and, if time allows, to investigate issues related to the mapping and connectivity of neurons in the system. We have obtained from a collaborator a silicon implementation of the elementary motion detectors in the fly, and the project would take this, together with a small motor platform and microcontroller, and build a complete robotic system using (and adding to) software we are currently developing for real-time robot control.

Involved inside group:

Involved outside group:

Reid Harrison, Department of Electrical and Computer Engineering, University of Utah

Recent publications:

Webb, B. and Reeve, R. Reafferent or redundant: How should a robot cricket use an optomotor reflex? Adaptive Behavior (2003), 11(3):137-158 (pdf)

Projects: