BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Spike-timing based neuronal information processing: applications t
 o vision and speech
DTSTART:20111201T151500
DTSTAMP:20260408T133401Z
UID:10534ff48d45269076e006d829c1f97a2203079b0c99164203e59a76
CATEGORIES:Miscellaneous
DESCRIPTION:Robert GÜTIG\, Racah Institute of Physics\, The Hebrew Univer
 sity\nThe timing of action potentials of sensory neurons contains substant
 ial information about the eliciting stimuli. Although computational advant
 ages of spike-timing-based neuronal codes have long been recognized\, it i
 s unclear whether and how neurons can learn to read out such representatio
 ns. We propose a novel biologically plausible supervised synaptic learning
  rule\, the tempotron\, enabling neurons to efficiently learn a broad rang
 e of decision rules\, even when information is embedded in the spatio-temp
 oral structure of spike patterns and not in mean firing rates. We demonstr
 ate the enhanced performance of the tempotron over the rate-based perceptr
 on in reading out spike patterns from retinal ganglion cell populations.\n
 \nExtending the tempotron to conductance-based voltage kinetics\, we show 
 that this model can subserve time-warp invariant processing of afferent sp
 ike patterns. Furthermore\, we show that the conductance-based tempotron c
 an learn to balance excitation and inhibition to match its integration tim
 e constant to the temporal scale of a given processing task. These mechani
 sms enable already small populations of model neurons to match the perform
 ance of state-of-the-art speech recognition systems on isolated word recog
 nition tasks. 
LOCATION:BC 01 https://plan.epfl.ch/?room==BC%2001
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR