BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Neuromorphic control - Luka Ribar\, University of Cambridge
DTSTART:20191114T140000Z
DTEND:20191114T150000Z
UID:TALK130111@talks.cam.ac.uk
CONTACT:Alberto Padoan
DESCRIPTION:The field of neuromorphic engineering aims to understand and m
 imic biological organisational principles in order to build hardware with 
 efficiency and performance of animal nervous systems. In this talk\, we wi
 ll look into the concept of neuromodulation as an important control mechan
 ism of neural systems and discuss how it can be replicated in simplified n
 euromorphic circuits. We will start by highlighting the essential feedback
  structure of excitable systems and show how complex neural dynamics and i
 ts modulation can be broken down into four basic feedback loops. We will t
 hen show how this feedback structure can be retained in a simplified circu
 it modelling framework. This architecture allows us to characterise the be
 haviour of the circuit through its input-output properties in the form of 
 I-V curves in a few distinct timescales. We will show how modulation of th
 e feedback loops shapes these I-V curves\, and in turn\, captures the tran
 sition between different neural oscillating regimes. The circuit structure
  can be easily synthesised using well-known neuromorphic building blocks w
 hich we will introduce and discuss in the talk. Finally\, we will consider
  how the control of neural excitability properties can be utilised in simp
 le network settings.
LOCATION:Cambridge University Engineering Department\, Seminar Room JDB
END:VEVENT
END:VCALENDAR