Neuromorphic architectures for real-time behaving systems

27 June 2013
Begin time: 
W0 - 135


For many practical tasks, conventional computing systems cannot match the performance of biological systems. One of the reasons is that the architecture of nervous systems, in which billions of nerve cells communicate with action potentials (so called "spikes") in parallel, is very different from that of today's computers. Recently developed brain-inspired hardware architectures that emulate the biophysics of neurons and synapses in silicon represent a promising technology for implementing alternative computing paradigms. In this presentation I will present an overview of past and present neurocomputing approaches, and  propose hybrid analog/digital circuits that directly emulate the properties of neurons and synapses. I will present a quick overview of the neuromorphic sensors being developed at the Institute of Neuroinformatics, University of Zurich and ETH Zurich, focus on circuit models of spiking neurons, synaptic dynamics, and synaptic plasticity mechanisms, and show how they can be configured to implement real-time compact neural processing systems. Finally, I'll propose neural architectures inspired by cortical circuits for implementing  artificial  real-time behaving neural systems.