Learning in Spiking and Deep Neural Networks implemented using Digital, Analog, and Memristive Devices

This talk will cover the design and implementation of a number of brain-inspired learning electronic circuits using different hardware approaches, including analog, digital, and memristors.

All these implementations are able to completely and with a minimal error replicate the outcome of a wide range of biological experiments. The talk will discuss the use of a number of these implementations in a spiking neural architecture comprising of different types of neurons to perform cognitive tasks such as pattern classification and unsupervised character recognition.

Furthermore, the talk will discuss results for the hardware implementation and simulation of non-spiking deep artificial neural networks on Field Programmable Gate Arrays (FPGA) and memristive crossbars. It also discusses the use of deep learning techniques for various applications ranging from agriculture to aquaculture and marine research at the Neural Inspired Computing and Engineering (NICE) Lab, James Cook University, Townsville, Australia.  

About the presenter

Mostafa Rahimi Azghadi completed his PhD in Electrical & Electronic Engineering at The University of Adelaide in July 2014, earning a Doctoral Research Medal as well as the 2015 Adelaide University Alumni Medal. He was a postdoctoral fellow at Adelaide and Sydney Universities between 2014 and 2016. In July 2016, Mostafa joined James Cook University, as a lecturer at the College of Science and Engineering, in Townsville. In 2018, Mostafa founded Neural Inspired Computing and Engineering (NICE) lab at JCU. He is currently an Associate Professor researching low-power and high-performance neuromorphic accelerators for neural inspired and deep learning networks for a variety of applications from agriculture to marine science. He has co-raised over $6.5M in research funding from national and international resources, over $1M of which as the lead investigator. 

A/Prof. Rahimi was a recipient of several national and international accolades including a 2015 South Australia Science Excellence award, a 2016 Endeavour Research Fellowship, a 2017 Queensland Young Tall Poppy Science Award, a 2018 JCU Rising Star ECR Leader Fellowship, a 2019 Fresh Science Queensland Finalist, and a 2020 JCU Award for Excellence in Innovation and Change. Dr Rahimi is a senior member of the IEEE, the chair of IEEE Northern Australia Section, and a TC member of Neural Systems and Applications of the circuit and system society. He serves as an associate editor of Frontiers in Neuromorphic Engineering and IEEE Access.