This was once a page for publications, I now rely on Google Scholar and ORCID to list all the important stuff. Get to Google Scholar by following this link► Other less important publications will be linked to on this page or under the appropriate project in the projects list.
Latest research is with the EU funded Blue Action project with the Arctic Centre (University of Lapland) building business models of a ski resort and analyzing weather forecasts and data from climate models.
I am continuing my collaboration with Reading University and Christchurch University Canterbury in educational research. The project is to develop workshops that have an impact on the way that school students engage with science, the type of people who are involved in it, and the type of problems it can (and cannot) help us with. Preliminary results were presented at BERA 2016 and the book will be out in Autumn 2017.
Much of my time in recent years has been devoted to MATLAB and C programming, computational studies of biological signal processing, and neuromorphic engineering. I have been developing novel approaches to feature extraction and recognition using statistical and information-theoretic approaches.
The latest paper co-authored with colleagues at Plymouth University and MTA Budapest is submitted.
I have recently posted a technical report ► suggesting ways that the techniques I have developed could be applied to image processing.
Older work
SCANDLE ► was an EU funded project that ran until spring 2012 that developed neuromorphic approaches to detection and classification of sound and behaviour.
EmCAP ► finished on a high in 2009. We concentrated in this project on the patterns and rhythms of music.
COLAMN finished in 2010 and was a project I only had a very small part in. It was concerned with the nature of cortical processing, a subject close to my heart (and head).
ALAVLSI ► was the project than funded my PhD. The link to the project website is broken and might never be restored. The project finished in 2005 and the servers hosting the site at ETH Zurich have since suffered a bit of a catastrophe.
The SCANDLE-mongers in Cyprus
The EmCAP-pers - or `mad dogs and Englishmen go out in the Budapest sun'.
Experiment that took 7 years to come to fruition included in Proceedings of ISH 2015 (to be published by Springer in November) investigating relative pitch perception.
Paper submitted to PLoS Comp. Biol. June 2015 suggest a new way that birds might extract features of conspecific song at a neural level - now under revision
I write computer programs, and help to design devices, that are "neuromorphic" - meaning they have the same form and work in the same way as the brain.
My principle research interest is to explore how we hear, and how we learn to hear. This has involved studies of musical perception, analysis of speech stimuli, and exploring hypotheses about how the neural substrate might extract meaningful features and landmarks from the auditory scene. To validate these ideas, they are then implemented in computer models that respect neurophysiological limitations, and in machines that are not digital or programmable but 'neuromorphic' - ie they contain silicon neurons and synapses.
Since 2000 I have been involved in a series of major international research projects exploring computational modelling of neural activity and it's relationship to neuromorphic engineering, behaviour, and measurements of brain activity. This has led to a series of papers in collaboration with colleagues at ETH (Zurich), MTA (Budapest), UPF (Barcelona) as well as in Plymouth. You can see a list my academic papers by clicking here.
My ideas are implemented as software simulations of neural activity. They are designed to throw light on how the brain works, but are also designed to be implemented in ground-breaking new hardware such as the multi-processor SpiNNaker boards (above left - follow this link for details) and the even more radical analogue neuromorphic chips being developed in Zurich (above right - follow this link for details).
In particular I investigate the way in which we process sound, from the ear to the brain. This is a complex and controversial issue that is not as well understood as many workers in the field would like to believe.
This work could be called "computational neuroscience". But because we also transfer our ideas in to silicon chips that incorporate silicon neurons it also falls under "neuromorphic engineering".