VIRTUAL EVENT: Making Siri More Like Us: The Neuromorphic Approach

As much as we love them, computers are quick to make us feel slow. Whether it’s Siri’s ability to calculate the square root of pi or that chess AI beating you more times than you would like to admit – it seems like we’ve lost the battle. But fortunately for us, perhaps not the war…

Your secret weapon is efficiency. You are capable of running the human body autonomously – an incredibly complex task – whilst also carrying out your day to day lives with your brain operating at the power consumption of a light bulb. And we’re not talking filament here, we’re talking your energy saving bulbs. Roughly.

In this talk we are going to explore the issue of power consumption within today’s machine learning hardware and discuss how researchers are taking inspiration from the human brain to tackle this issue. Welcome to neuromorphic engineering.

Speaker Bio

Dan Mannion is a PhD candidate and studies the use of novel electronic devices in bio-inspired computing. Based in the department of electronic and electrical engineering at UCL he has studied the field of neuromorphic engineering for the last 5 years. The long-term goal of this research is to reduce the power consumption of machine learning hardware by taking inspiration from our very own brains. Recently, he developed circuits to implement edge detection of an image in real time with minimal computation – awarded the NEUROTECH prize.

He also collaborates with Birkbeck University working to conserve pieces of art and cultural heritage. In this role he has worked with the Tate Britain and the Mary Rose Trust.

Twitter: @daniel_mannion