Paving The Way For Next-generation AI

The Department of Electronics and Communication Engineering is organising an interaction for its faculty with the acclaimed academician, Dr Bharat Ramesh, Senior Lecturer at International Centre for Neuromorphic Systems, Western Sydney University, Australia, on May 8, 2023. Dr Ramesh will speak on Neuromorphic Engineering – Paving the way for Next-generation AI.

Abstract of the talk

Neuromorphic Engineering is a subfield of Electrical Engineering that aims to apply knowledge of how signals are processed in the brain to build electronic signal processing systems that vastly outperform current digital signal processing systems. Current ‘smart’ sensors are generally sensors with a built-in computer. We aim to develop smart sensors with a built-in brain by combining bio-inspired sensors with bio-inspired signal processing.

Reverse Engineering the Brain – The brain creates a coherent interpretation of the external world based on input from its senses. Yet data from the senses are unreliable and confused. How does the brain determine what is out there in the world around it? Neuromorphic Engineering conducts neurophysiological and psychophysical investigations combined with theoretical, computational, and electronic modelling studies to discover how the brain achieves this. The outcomes of this research will then be applied to create intelligent electronic sensors.

This talk reviews the fundamentals and properties of neuromorphic sensors and provides a unified view of the main approach taken. Examples are presented to demonstrate the application using a novel setup known as Astrosite. Only about 40% of those satellites are active, so the potential for disruptions caused by dead satellites and debris in orbit that don’t perform as expected has increased the need to track and monitor objects in space. Western Sydney University has been tackling this problem with bio-inspired (neuromorphic) engineering principles. Using unique event-based sensors on mobile telescopes, they have demonstrated the ability to capture and track real-time orbital objects day and night.

About the Speaker

Dr Bharath’s primary research interests include pattern recognition and computer vision. His research is centred on event-based cameras for computationally efficient machine learning for problems such as object recognition and related areas such as scene understanding, face recognition, and object detection. He has published and reviewed papers at well-known AI/ML conferences. He also serves as a reviewer for top conferences such as the CVPR and journals like Transactions on Image Processing (TIP). He received a BE degree in electrical & electronics engineering from Anna University of India in 2009; an MSc and PhD in electrical engineering from the National University of Singapore in 2011 and 2015, respectively, while working at the Control and Simulation Laboratory on Image Classification using Invariant Features.

To learn more about this cutting-edge technology, join the session!