Opportunities
Research Associate in Computational Statistics for Image Data
Duration: 12 months in the first instanceSalary: tbc
Start date: tbc
Closing date: tbc
Vacancy Information
Applications are invited for a postdoctoral research associate position in the Department of Statistical Science at University College London, UK. This DSTLfunded project will draw upon exciting recent developments in applied and computational statistics and beyond to advance current understanding of how relational and contextual aspects of image data can contribute to tasks such as detection, classification, and segmentation.Description
The project will address the current need to progress new spatialstatistical models and computeraided approaches that have the capability to capture both the rich statistical, and structural, interdependencies present in highdimensional image data sets. The work will draw on the interacting and convergent fields of computational statistics, machine learning, statistical signal processing, and computer vision where recent activity has given rise to a wealth of concepts that attempt to describe both correlatory and relational interactions of heterogeneous data over, and between, multiple scales. Image data sets taken from a selection of neoteric applications, including the life sciences, astronomy, and oceanography, will be considered.Of some interest to this project is the use and development of graphs, random fields and point processes, stochastic geometry, hierarchical representations, and/or multiresolution analyses to model the dependencies between key constituent parts of image data. It is hoped that this somewhat more comprehensive, relational, description of the data will facilitate better, and more robust, detection and segmentation. The postholder will have an interest in one or more of these (or allied) areas along with experience of some of the associated inferential/estimation machinery (such as Markov chain Monte Carlo, lasso, etc, and variants thereof). Good programming skills in Matlab, R, Python, and/or C/C++ are essential.
It is anticipated that the project will afford the opportunity for the postholder to enjoy some collaboration with Dstl, industry, other universities in the UK (such as Cambridge, Bristol, and Imperial) and elsewhere, and multiple departments and centres at UCL, including the Centre for Computational Statistics and Machine Learning. In particular, there will be ample opportunities for interaction with the immediate research group which currently comprises: one research associate and three Phd students also funded by Dstl; one EPSRCfunded research associate; and two other Phd students funded by an ASTAR/UCL jointscheme.
The post comes with a generous travel and research expenses budget and is funded for 12 months in the first instance.
The Department of Statistical Science at UCL offers a vibrant and intellectually stimulating environment for individuals who wish to develop their careers in a worldclass research environment. UCL is ranked among the top ten research institutions worldwide and has unique strengths in Computational Statistics and Machine Learning. Together with the Gatsby Institute for Computational Neuroscience at UCL, the Departments of Statistical Science and Computer Science form the Centre for Computational Statistics and Machine Learning which is part of the European network PASCAL.
Informal enquiries to Dr James Nelson are welcomed.
How to apply
Instructions will appear here soon.Recent grants
 Dstl RA project on "Graphbased, informationtheoretic
approaches to minimal sensing" (2014)
 Dstl Impact PhD studentship "Regularity and
information filtering approaches to uncertainty management"
(2014)
 EPSRC/D2U RA project on "Semisupervised learning and
smart sensor network monitoring" (2014)
 ASTAR/UCL joint PhD studentship "Statistical analysis,
semantic modelling, and learning for highly structured data"
(2014)
 ASTAR/UCL joint PhD studentship "Multiresolution
hypergraphs for histopathological classification" (2013)
 Dstl Impact PhD studentship "Inhomogeneous multiresolution
random fields" (2013)
 EPSRC/TSB Technology Inspired Innovation RA project on
"Smart sensor network for powerline monitoring using machine
learning" (2013)
 Dstl fully funded 4 year PhD studentship "Statistical
signal processing and machine learning for network traffic
anomaly detection" (2012)
 Dstl/Atlas Elektronik UK RA project on semisupervised
mine countermeasures (2012)
 EPSRC RA project "Multiresolution Markov models for detecting radial patterns of spicules in mammograms" (2012)

EPSRC RA project "Development of locally invariant signal
processing to discriminate between key manmade and natural
features" (2009)
 MoD Counter Terrorism Centre project "Anomaly detection in video imagery" (2009)
Principal Investigator
Researcher CoInvestigator
Recent consultancy
 2011: Systems Engineering & Assessment Ltd
 2011: Defence Science and Technology Laboratory
Publications
 Tomassi, D., Milone, D., and Nelson, J. D. B. (2014+)
"Wavelet shrinkage using adaptive structured sparsity
constraints" Signal Processing (in press)
 Nelson, J. D. B. (2014) "On the equivalence between a minimal codomain cardinality Riesz basis, a system of HadamardSylvester operators, and a class of sparse, binary optimisation problems" IEEE Transactions on Signal Processing (in press)
 Gibberd, A. J. and Nelson, J. D. B. (2014) "High
dimensional changepoint detection with a dynamic graphical
lasso" IEEE International Conference on Acoustics, Speech,
and Signal Processing
 Nafornita, C., Isar, I., and Nelson, J. D. B. (2014)
"Regularised, semilocal Hurst estimation via generalised lasso
and dualtree complex wavelets" IEEE International
Conference on Image Processing
 Nelson, J. D. B. and Krylov, V. (2014) "Textural lacunarity for
semisupervised learning in sonar imagery" IET Radar, Sonar
& Navigation 8(6):616621
 Kalaitzis, A. and Nelson, J. D. B. (2014) "Smoothed, grouped
sparse coding for anomaly detection" IEEE Machine Learning for
Signal Processing
 Nelson, J. D. B. (2013) "Fused Lasso and rotation
invariant autoregressive models for texture classification"
Pattern Recognition Letters 34(16):21662172
 Krylov, V. and Nelson, J. D. B. (2013) "Fast road network
extraction from remotely sensed images" Advanced Concepts for
Intelligent Vision Systems, Lecture Notes in Computer Science
8192:227237
 Krylov, V., Taylor, S., and Nelson, J. D. B. (2013) "Stochastic extraction of elongated curvilinear structures in mammographic images". International Conference on Image Analysis and Recognition
 Gibberd, A. J. and Nelson, J. D. B. (2013) "Sparsity for
change point detection in network traffic". Data Analysis
for CyberSecurity

Nelson, J. D. B. and Kingsbury, N. G. (2012) "Fractal
dimension, wavelet shrinkage, and anomaly detection for
mine hunting". IET Signal Processing Journal
 Nelson, J. D. B. and Kingsbury, N. G. (2012) "Multiresolution Markov random field wavelet shrinkage for ripple suppression in sonar imagery". 9th IMA International Conference on Mathematics in Signal Processing

Julier, S., De Nardi, R., and Nelson, J. D. B. (2012)
"Multirate estimation of coloured noise models in graphbased
estimation algorithms". International Conference on
Information Fusion

Nelson, J. D. B. and Kingsbury, N. G. (2011) "Enhanced shift
and scale tolerance for dualtree complex wavelet rotation
invariant matching". IEEE Transactions on Image
Processing, 20(3):814821

Nelson, J. D. B. and Kingsbury, N. G. (2010) "Fractal
dimension based sand ripple suppression for mine hunting
with sidescan sonar". International Conference on
Synthetic Aperture Sonar and Synthetic Aperture Radar

Nelson, J. D. B. and Kingsbury, N. G. (2010) "Dualtree
wavelets for locally varying and anisotropic fractal
dimension estimation". IEEE International
Conference on Image Processing

Nelson, J. D. B., Damper, R. I., Gunn, S. R. and Guo, B.
(2009) "A signal theory approach to support vector
classification: the Sinc kernel". Neural Networks 22
(1): 4957

Pang, S. K., Nelson, J. D. B., Godsill, S. J., and Kingsbury,
N. G. (2009) "Video tracking using dualtree wavelet polar
matching and RaoBlackwellised particle filter". Journal
on Advances in Signal Processing

Nelson, J. D. B., Damper, R. I., Gunn, S. R. and Guo, B.
(2008) "Signal theory for SVM kernel design with applications
to parameter estimation and sequence kernels"
Neurocomputing 72 (13): 1522

Guo, B., Gunn, S. R., Damper, R. I. and Nelson, J. D. B.
(2008) "A fast separabilitybased feature selection method for
highdimensional remotelysensed image classification".
Pattern Recognition 41 (8): 16701679

Guo, B., Gunn, S. R., Damper, R. I. and Nelson, J. D. B.
(2008) "Customizing kernel functions for SVMbased
hyperspectral image classification" IEEE Transactions on
Image Processing 17 (4): 622629

Nelson, J. D. B., Pang, S. K., Godsill, S. J., and Kingsbury,
N. G. (2008) "Tracking ground based targets in aerial video
with dualtree complex wavelet polar matching and particle
filtering". International Conference on Information
Fusion

Pang, S. K., Nelson, J. D. B., Godsill, S. J., and Kingsbury,
N. G. (2008) "Video tracking using dualtree wavelet polar
matching and particle filtering". The Institution of
Engineering and Technology Seminar on Target Tracking and Data
Fusion

Mahmood A., Tudor, P. M., Oxford, W., Hansford, R., Nelson J.
D. B., Kingsbury, N. G., Katartzis, A., Petrou, M.,
Mitianoudis, N., Stathaki, T., Achim, A., Bull, D.,
Canagarajah, N., Nikolov, S., Loza, A., and Cvejic, N. (2007)
"Applied multidimensional fusion". The Computer Journal
50 (6)

Guo, B., Gunn, S. R., Damper, R. I. and Nelson, J. D. B.
(2006) "Band selection for hyperspectral image classification
using mutual information". IEEE Geoscience and Remote
Sensing Letters

Nelson, J. D. B., Damper, R. I., Gunn, S. R. and Guo, B.
(2006) "Signal theory for SVM kernel parameter
estimation". Proceedings of IEEE International
Workshop on Machine Learning for Signal Processing

Nelson, J. D. B. (2005) “A wavelet filter enhancement
scheme with a fast integral Bwavelet transform and pyramidal
multiB wavelet algorithm”. Applied &
Computational Harmonic Analysis 18(3): 234251

Guo, B., Gunn, S. R., Damper, R. I. and Nelson, J. D. B.
(2005) "Adaptive band selection for hyperspectral image fusion
using mutual information". International Conference on
Information Fusion

Guo, B., Gunn, S. R., Damper, R. I. and Nelson, J. D. B.
(2005) "Hyperspectral image fusion using spectrally
weighted kernels". International Conference on
Information Fusion

Nelson, J. D. B. and Fu, S. (2004) "A yaw and tilt invariant
vehicluar egoposition model". Technical Report,
Cranfield University

Nelson, J. D. B. (2003) "On the coefficient quantization of
the Fourier basis". IEEE Transactions on Signal
Processing 51(7): 1838–1846

Nelson, J. D. B. (2003) "A multichannel, multisampling
rate theorem". Sampling Theory in Signal and Image
Processing: An International Journal 2(1): 83–96
 Nelson, J. D. B. (2003) "Applications of
enhanced vision to the automotive industry". SMi
Conference on Enhanced Vision

Nelson, J. D. B. (2001) "The construction of some Riesz basis
families and their application to coefficient quantization,
sampling theory, and wavelet analysis" PhD thesis, Anglia
Polytechnic University
Selected Invited Talks and Events
 Gibberd, A. J. and Nelson, J. D. B. (2014) "Graphical
statistics for anomaly detection in large dynamic datasets"
Defence Science and Technology Laboratory Statistics
Interest Group Forum
 Le Roux, J. A. and Nelson, J. D. B. (2013) "Integrated variance
estimation of highfrequency data in the presence of
microstructure noise" Young Statisticians Meeting
 Gibberd, A. J. and Nelson, J. D. B. (2013) "Catching
cyberattacks with a LASSO" Young Statisticians Meeting
 2012 U.S. Office of Naval Research Intersections in
Signal Processing, Acoustics, and Automatic Target Recognition for
Maritime Applications, Washington
 2012 EPSRC/DST Interaction Meeting in Applied Mathematical
Sciences Research Challenges, Edinburgh
 Nelson, J. D. B. and Kingsbury, N. G. (2011)
"Fractal dimension, wavelet shrinkage, and anomaly detection for
mine hunting in sonar imagery". University Defence Research
Centre Industry day, Imperial College London
 Nelson, J. D. B. and Kingsbury, N. G. (2010) "Estimation of
fractal dimension and shiftscale tolerant, rotation invariant,
matching". University of Surrey
 Nelson, J. D. B. and Kingsbury, N. G. (2010) "Dualtree
wavelets for locally varying and anisotropic fractal dimension
estimation". Edinburgh Research Partnership in
Engineering and Mathematics: Signal and Image Processing
Joint Seminar Programme, Edinburgh University and
HeriotWatt University
 Kingsbury, N. G. and Nelson, J. D. B. (2009) "Dualtree
wavelets for fractal dimension estimation". University
Defence Research Centre Launch, Imperial College London
 Nelson, J. D. B. and Kingsbury, N. G. (2009) "Detection and
displacement estimation with the DTCWT". Defence Science
and Technology Laboratory Maths Interest Group Forum
 Nelson, J. D. B. and Kingsbury, N. G. (2009) "Detection and
displacement estimation with the DTCWT". Defence Science
and Technology Laboratory Research Exploitation Day
 Nelson, J.D.B., Chen, H., Kingsbury, N. G. (2009) "3D MRI
registration with dualtree wavelets". Annual General
Meeting, The Magnetic Resonance Group of the Institute of
Physics
Teaching
 2010  2011, 2013  present: Lecturer of
"LTCC course on Stochastic Processes" (London Taught Course
Centre: an EPSRC supported graduate training programme for PhD
students in the mathematical sciences)
 2011  present: MSc "Statistical Computing"
 2011  present: Tutor for "Introduction to Applied Probability"
 2010  2011: Tutor for MSc "Skills Development course in
graphics"
 2011: Lecturer for the MSc Foundation course on "Poisson
Processes & Further Applied Probability"
 2010  2011: Lecturer of "Forecasting"
 2010  2011: Tutor for "Probability and Inference"
 2010: Tutor for "Introduction to probability and
statistics"
Admin
 Natural
Sciences Mathematics and Statistics stream representative
 Centre for Security Technology steering committee member
 Statistical
Science Departmental Teaching Committee member
Code

Construction of minimal codomain cardinality Riesz bases
cf.: Nelson, J. D. B. (2014) "On the equivalence between a
minimal codomain cardinality Riesz basis, a system of
HadamardSylvester operators, and a class of sparse, binary
optimisation problems" IEEE Transactions on Signal Processing
(in press)

MCMC extraction of curvilinear structures in mammogram data
cf.: Krylov, V., Taylor, S., and Nelson, J. D. B. (2013)
"Stochastic extraction of elongated curvilinear structures in
mammographic images". International
Conference on Image Analysis and Recognition