Email: j.dymond@soton.ac.uk
My research focuses on energy efficient deep learning applications, specifically I investigate approaches that aim to maximally utilise resources available to them. Applications of this research range from search and rescue to mobile implementations.
My current research focusses on a formulation of machine learning called 'Progressive Intelligence'. This is a type machine learning which adapts to resource availability at runtime. It achieves this by processing inputs incrementally, first producing low-cost--lower-confidence outputs, and improving upon these incrementally.
My work has centred around developing progressive intelligence systems, most recently adapting branched neural networks to present progressive intelligence properties.
Graduated from the University of Sheffield in 2019 with an MPhys in Physics and Astrophysics. Now undertaking a PhD in resource efficient machine learning, as part of the MINDS CDT at the University of Southampton.