The CAMERA Expert Speaker Series offers free, regular seminars which are open to all. The series brings leading experts from both industry and academia to the University of Bath.
The Expert Speaker Series supports CAMERA's work to accelerate the real world impact of fundamental research across multiple disciplines. Our speaker series will showcase knowledge spanning our core research themes of creative technologies, rehabilitation and human performance to a wider audience.
For those unable to attend, please see individual speaker information for seminar video.
Due to anticipated high demand, all events require a ticket for entry. Please see details for each event to apply for a ticket. Each seminar will be followed by a drinks reception.
Professor Taku Komura, Edinburgh University
Neural State Machine for Character-Scene Interactions
2pm Thur 4th June 2020, remote
Taku will cover recent developments in neural network-based character controllers. Using neural networks for character controllers significantly increases the scalability of the system – the controller can be trained with a large amount of motion capture data while the run-time memory can be kept low. As a result, such controllers are suitable for real-time applications such as computer games and virtual reality systems.
The main challenge is in designing an architecture that can produce movements in production-quality and also manage a wide variation of motion classes. Our development covers low level locomotion controllers for bipeds and quadrupeds, which allow the characters to walk, run, side-step and climb over uneven terrain, as well as a high level character controller for humanoid characters to interact with objects and the environment, which allows the character to sit on chairs, open doors and carry objects. In the end of the talk, I will discuss about the open problems and future directions of character animation.
Taku Komura is a Professor at the Institute of Perception, Action and Behaviour, School of Informatics, University of Edinburgh. As the leader of the Computer Graphics and Visualization Unit his research has focused on data-driven character animation, physically-based character animation, crowd simulation, cloth animation, anatomy-based modelling, and robotics.
Recently, his main research interests have been the application of machine learning techniques for animation synthesis. He received the Royal Society Industry Fellowship (2014) and the Google AR/VR Research Award (2017).
No recording is available of this seminar.
Professor Alan Wilson, Royal Veterinary College
Anatomical and mechanical constraints to athletic performance: studies in cheetahs, horses, humans and other animals.
4pm, Tues 7th May, CB 2.6, University of Bath
Alan will talk about the specialised anatomy of athletic animals, with particular reference to muscle-tendon interaction. He will talk about the development of technology using high accuracy GPS and inertial sensors for studying locomotion of free-ranging animals in their natural environment and processing of those data. He will discuss the insights gained from using his novel technology in studies of African predators and their prey, free flying birds and occasionally humans. These include the costs and benefits of birds flying in a flock, the tactics an antelope should use to evade capture by a cheetah and the mechanisms that enable a wildebeest to walk 80 km over five days in 40oC without drinking.
Alan Wilson is Professor of Locomotor Biomechanics and head of the Structure and Motion Lab at the Royal Veterinary College. www.rvc.ac.uk/sml
He holds first degrees in veterinary science and physiology, a PhD in tendon biomechanics and has worked in animal locomotion research since qualifying in 1987. He is registered as a Veterinary Surgeon in Botswana and South Africa specialising in wildlife capture and he is the main pilot for the RVC research aircraft.
His research interests include the design of animals for high speed and economical locomotion, innovative measurement techniques for studying animals during field locomotion, and muscle-tendon interaction in locomotion. He has held BBSRC, EPSRC, ERC, CHDI (charity) and DARPA (US Defence Agency) funding to study locomotor biomechanics in a range of species. He has held three BBSRC CASE awards, developing GPS based localisation technology (for terrestrial and aerial locomotion), integrating physiological data with ultra-wideband radio derived speed and position data (for horseracing), and applying wireless sensors for physiological measurements respectively. His DARPA funding with Boston Dynamics explored design principles of cheetah- and dog- inspired high-speed robots. He has published papers demonstrating and validating innovative techniques for measurement of function during high speed locomotion and papers using these techniques to demonstrate basic biological mechanisms (including eight in Nature and one in Science). His google scholar h-index is 53 and 97 of his papers have received more than ten citations in the last five years.
Dr Will Smith, University of York
What can vision and graphics learn from physics?
4pm, Weds 30th January 2019, Weston Studio, The Edge
In the past 6 years, machine learning has transformed computer vision. This learning-based approach treats computer vision as a problem of finding a function that approximately maps directly from images to desired output. The success of this approach has been driven by the availability of very large, labelled datasets, advances in the training and architecture of deep neural networks and developments in GPU hardware. Recent work, particularly using generative networks, shows that learning-based approaches could have a similar impact in computer graphics.
This talk will argue that learning alone, specifically supervised learning, cannot be used to solve all problems in visual computing. Using example problems in which training data is either very limited or does not exist at all, I will show how models borrowed from physics (with a bit of statistics and geometry thrown in) can be used to supervise learning, i.e. learning of a task can be “self-supervised” by explicit models. Specifically, I will show recent results in nonlinear 3D shape modelling, inverse rendering in the wild and biophysical face image interpretation that apply this philosophy.
Dr Will Smith is a Senior Lecturer in the Computer Vision and Pattern Recognition group in the Department of Computer Science at the University of York. He completed a BSc and PhD in Computer Science, both at the University of York, in 2002 and 2007 respectively. He has published more than 100 papers in international journals and conferences.
His areas of interest include 3D face capture, illumination and reflectance modelling, face processing, statistical shape modelling (particularly for directional data or data lying on non-Euclidian manifolds), psychology and neuropsychology of face perception, modelling craniofacial variation, omnidirectional imaging, shape-from-shading and photometric stereo.
Dr Raj Sengupta, The Royal National Hospital for Rheumatic Diseases, Bath (RNRHD)
Inflammatory Arthritis, Machine Learning and Digital Technologies: the future of diagnosis and management.
5pm, Tues 13th November 2018, Weston Studio, The Edge
The diagnosis and management of rheumatic diseases has evolved considerably over the last 2 decades. We are also witnessing a rapid progress of new technological advances in healthcare. The development and validation of new technologies in rheumatic disease provides an exciting opportunity to evolve and optimise personalised care for your patients in the 21st century. This talk will describe current rheumatology practice and explore opportunities to develop new technologies in the field.
Dr Raj Sengupta is a Consultant Rheumatologist and Lead for Ankylosing Spondylitis at The Royal National Hospital for Rheumatic Diseases, Bath (RNRHD). He is a member of the BSR Spondyloarthritis Special Interest Group, the BSR AxSpA Biologics Guidelines Committee and a member of ASAS (Assessment of Spondyloarthritis International Society).
He has represented the BSR at the NICE TA383 appraisal for anti TNF in AxSpA. Dr Sengupta is a trustee and medical advisor for the National Ankylosing Spondylitis Society (NASS) and is one of the founder members of BRITSpA. Dr Sengupta was the recipient of the 2016 NASS Patients Choice award for the Best care provided by a Rheumatologist.
He is the principal investigator in several national and international clinical trials in Axial Spondylarthritis. He has a number of peer reviewed publications on AxSpA. Dr Sengupta has been an invited speaker on AxSpA at several national and international meetings
Professor Richard Bowden, University of Surrey
Vision and AI: then, now and tomorrow
Tuesday 25th September, 6pm- 8pm, Weston Studio, The Edge, University of Bath
Computer vision broke away from AI as a field in its own right in the 1960’s. As a fundamental human sense “vision” is something we take for granted, but the power of human visual ability lies in the fact that it is implicitly tied to our knowledge of the world and our ability to reason. Computer vision is now a huge field, with major investment from industry. Who could have foreseen that every one of us would carry at least one device with a camera all the time. For many tasks, machine vision can surpass human performance but these areas tend to be quite specific in domain. We don’t yet have a generic visual machine that can adapt to new tasks the way a human can, and yet we are expecting SAE level 5 fully autonomous vehicles to be on our roads within the next 10 years.
This talk tries to summarise the field as it has developed. It will look at two ongoing research problems at Surrey, that of sign language translation and autonomous vehicles. Where we are now and what problems remain to be solved. In the context of the field we will discuss the shortcomings of current AI and what work needs to be done to truly achieve human performance.
Richard Bowden is Professor of Computer Vision and Machine Learning at the University of Surrey where he leads the Cognitive Vision Group within CVSSP and is Associate Dean for postgraduate research within his faculty. His research centres on the use of computer vision to locate, track, understand and learn from humans. He has held over 40 research grants from UK, EU funding bodies as well as industrial funded projects. These projects cover areas such as cognitive robotics and vision, sign and gesture recognition, lip-reading and nonverbal communication as well as many fundamental topics to computer vision such as tracking and detection. His research has been recognised by prizes, plenary talks and media/press coverage including the Sullivan thesis prize in 2000 and many best paper awards.
To date, he has published over 190 peer reviewed publications and has served as either program committee member or area chair for ICCV, CVPR, ECCV, BMVA, FG and ICPR in addition to numerous international workshops and conferences. He is an Associate Editor for the journals Image and Vision Computing and IEEE Trans Pattern Analysis and Machine Intelligence (the top journal in his field). He was awarded a Royal Society Leverhulme Trust Senior Research Fellowship in 2013 and is a member of the RS international exchanges committee. He was a member of the British Machine Vision Association (BMVA) executive committee and a company director for seven years. He is a member of the BMVA, a senior member of the IEEE and a Fellow of the Higher Education Academy. He was awarded a prestigious Fellowship of the International Association of Pattern Recognition in 2016.