Giacomo Indiveri

Wednesday 6th June 2018

Time:4.00pm

Ground Floor Seminar Room

25 Howland Street, London, W1T 4JG

Neuromorphic electronic circuits for building cognitive agents able to carry out real-time sensory processing and context-dependent computation

Artificial computing systems are vastly outperformed by biological neural processing ones for many practical tasks that involve sensory perception and real-time interactions with the environment, especially when size and energy consumption are factored in. One of the reasons is that biological neural processing systems, which comprise billions of neurons that communicate in parallel mainly via asynchronous action potentials, are very different from today's artificial computing systems, based mainly on serial communication and synchronous logic. Recent machine learning algorithms have taken inspiration from the nervous system to develop neuro-computing algorithms that are showing state-of-the-art performance in pattern recognition tasks. In parallel, different types of brain-inspired hardware architectures are being developed that reproduce some of the principles of computation used by the nervous system. These architectures represent a promising technology for both implementing the latest generation of neural networks, and for building faithful models of biological neural processing systems. In this talk I will present examples of electronic circuits that can be used to directly emulate the dynamics of real neurons and synapses, and showcase spike-based neural network architectures that can be used to perform neural computation, signal processing, and pattern recognition in real-time. I will show examples of VLSI neuromorphic processors developed in our Institute and present examples of spike-based learning, decision making, sensory processing and state-dependent computation.

Biography:

Giacomo Indiveri is a Professor at the Faculty of Science of the University of Zurich, Switzerland, and director of the Institute of Neuroinformatics of the University of Zurich and ETH Zurich. He obtained an M.Sc. degree in electrical engineering and a Ph.D. degree in computer science from the University of Genoa, Italy. He was a post-doctoral research fellow in the Division of Biology at Caltech and at the Institute of Neuroinformatics of the University of Zurich and ETH Zurich. He holds a "habilitation" in Neuromorphic Engineering at the ETH Zurich Department of Information Technology and Electrical Engineering. He was awarded an ERC Starting Grant on "Neuromorphic processors" in 2011 and an ERC Consolidator Grant on neuromorphic cognitive agents in 2016. His research interests lie in the study of neural computation, with a particular focus on spike-based learning and selective attention mechanisms. His research and development activities focus on the full custom hardware implementation of real-time sensory-motor systems using analog/digital neuromorphic circuits and emerging VLSI technologies.