Photo of Dr Jacqueline Christmas

Dr Jacqueline Christmas

Lecturer in Computer Science

Email:

Telephone: 01392 723039

Extension: (Streatham) 3039

Research interests

Machine learning for intelligent image and video understanding.  Bayesian modelling and variational approximation.  Sea wave prediction and quiescent period prediction.
 

Current projects

Sea Wave Prediction
Working with Prof. Michael Belmont to make significant improvements to the efficiency of sea wave energy converters by providing them with very short-term predictions of the profiles of the waves they are about to encounter. Based on this we are pioneering research into Quiescent Period Prediction (QPP) which aims to predict when short periods of relative calm are about to occur. This has the potential to allow a range of wave critical marine operations to be safely carried out at considerably larger wave amplitudes than would otherwise be possible. The models currently undergoing sea trials are currently deterministic in nature. I am in the process of developing a range of statistical models that are robust to the outliers we have seen in real sea data and able to suppress waves components for which the evidence is low.

Photogrammetry Reflectance Transformation Imaging
Working with Dr Judith Bannerman and Prof. George Bevan to develop new techniques in RTI and its use for photogrammetry.  A new web page for this work is coming soon.
 

Exeter Imaging Network

I am one of the organisers of the Exeter Imaging Network (EIN), which brings together researchers across all disciplines who use images in their work.  Our seminars occur approximately bi-monthly, with talks from invited speakers and space to discuss opportunities for cross-pollination of research ideas and to form new collaborations.
 

Previous projects

Using variational Bayesian models to

  • track wild crickets (Gryllus campestris) from recorded video clips.  The crickets are labelled with alphanumeric tags which we aim to recognise and we would like to identify certain behaviour patterns.  In collaboration with Prof. Tom Tregenza and Prof. Richard Everson.

  • identify cartilage cells in 3D confocal microscopy images and then match the cells in images taken before and after a mechanical stretch is applied to the cartilage sample.  In collaboration with Dr James Bell and Prof. Peter Winlove.

  • (see "Robust spatio-temporal latent variable models ") model EEG data and to estimate missing values in satellite images of phytoplankton in the surface of the sea.


 

Google+