September’s meeting is looking at the use of open source in state of the art AI.

AGENDA
18:00 – Feel free to join the online meeting to chat with other participants (tea and coffee for physical attendees).
18:30 – Presentations
20:00 – Closing discussion

This will be a hybrid meeting, with some people attending in person in London and others able to join via videoconference.

Attending in person

For those wishing to attend in person, due to COVID-19 limits, REGISTRATION IS ESSENTIAL – without a ticket, you will not be able to get in.  Numbers attending will be restricted, and there will be a waiting list.

Register to attend in person

NOTE. If you register and then find you can’t attend, please be sure to cancel, so we can offer the place to someone on the waiting list.

Attending online

For remote attendees, there is no requirement to register, you can just connect to the videoconference using BigBlueButton using this link.

NOTE. The videoconference link has changed since this event was first published.

We are also recording the talks for later posting on our YouTube channel.

The livestream link will be open from 18:00 for networking, and the event will start at 18:30 prompt. We’ll keep the link open afterwards for discussion.

Not a BCS member? Don’t worry! Our monthly newsletter provides the latest news, information and events in the world of IT. We’re here to keep you updated! Sign up today at campaign.bcs.org/bcs-supporter.  If you would just like to be added to the friends of the Open Source SG mailing list, then please send an email to ossgnonmem@lists.bcs.uk with subscribe in the subject.  You can unsubscribe at any time by emailing unsubscribe-ossgnonmem@lists.bcs.uk.

The videos are available here.

An open-source framework for multimodal artificial intelligence

Adrian Hopgood, University of Portsmouth

A wide range of techniques has emerged from the field of AI including neural networks, deep learning, rules, frames, model-based reasoning, case-based reasoning, Bayesian updating, fuzzy logic, multiagent systems, swarm intelligence, and genetic algorithms. They are all ingenious and useful in narrow contexts. It will be argued in this presentation that a truly intelligent system needs to draw on a variety of these approaches within a hybrid system. An open-source multiagent software framework called DARBS (Distributed Algorithmic and Rule-based Blackboard System) is proposed for this purpose. Several practical examples will be presented, ranging from image interpretation to the control of specialised manufacturing processes.

Adrian Hopgood is Professor of Intelligent Systems at the University of Portsmouth, where he is Director of Future & Emerging Technologies and of the South Coast Centre of Excellence in Satellite Applications. He is a Chartered Engineer, Fellow of the BCS (the Chartered Institute for IT), and a committee member for the BCS Specialist Group on Artificial Intelligence. He has worked at the level of Dean and Pro Vice-Chancellor in four universities in the UK and overseas, and has enjoyed scientific roles with Systems Designers PLC and the Telstra Research Laboratories in Australia. His main research interests are in AI and its practical applications. He has supervised 19 PhD projects to completion and published more than 100 research articles. His text book “Intelligent Systems for Engineers and Scientists: A Practical Guide to Artificial Intelligence” is ranked as a bestseller and its fourth edition is due in December 2021.

Open-source tools in machine learning applied to medical imaging: research needs and regulatory perspective

Sara Lorio & Jo Hobbs, LifeHub Bayer UK

Today’s clinical routine generates a vast amount of radiology data that needs to be assessed by healthcare professionals. Artificial intelligence (AI) can leverage this data in order to help with the increasing workload and improve the extraction of quantitative knowledge from the rich information present in medical images. In order to develop reliable and robust AI algorithms, it is crucial to process the imaging data and to evaluate the performance of the algorithm on multiple data sources. In this talk, we will cover the necessary steps to prepare medical images for the development of AI algorithms and we will talk about the regulation of AI technologies for clinical practice.

Sara Lorio is the technical lead at LifeHub Bayer UK.  She leads research projects for algorithm development in digital diagnostics and medical imaging. She is very passionate about combining different imaging techniques with cutting-edge data analysis in order to improve disease diagnosis and therapy.

Prior to this role, Sara held research positions at the University College of London and King’s College London, UK in collaboration with London main research hospitals. She received her PhD from University of Lausanne, Switzerland. Her research interests are in Medical Physics, Medical Imaging, Machine Learning and Neurobiology.

Jo Hobbs is the project support manager for the LifeHub UK. In this role she is supporting on a range of AI related projects being carried out within the LifeHub. She has a strong interest in how we can use innovative new technology, combined with what we already know, to promote better health care for everyone.

Open Source Dynamic Causal Modelling of COVID-19

Will Jones, Embecosm

Dynamic Causal Modelling is a state of the art AI modelling technique that reverse engineers an observed time series into a set of causal components and relationships. DCM has historically been developed for and applied to problems in neuroscience and brain imaging, but the technique is a very general one that has more recently, for example, been applied to modelling the COVID-19 pandemic with excellent results.

The standard implementation of DCM is open source, but it’s current implementation is in MATLAB, a proprietary tool. In this talk I discuss my work on creating a standalone implementation of one particularly application of Dynamic Causal Modelling (that of COVID-19) compatible with the open source GNU Octave language.

Dr Will Jones is head of AI and Machine Learning for Embecosm.  He recently completed his PhD at the University of Kent, which can be simply summarized as attempting to create a rigorous mathematical framework for the definition of artificial consciousness.