Advanced Digital Signal Processing

12-16 February 2007



This intensive course is designed to give an appreciation of the theory and application of digital signal processing. Starting at introductory level, the course covers fundamental theory including the Discrete Fourier Transform, and Fast Fourier Transform algorithms; it then progresses into the design of digital filters.

The theory and applications of adaptive filters is covered, and the course then moves into a discussion of statistical signal processing, including Bayes estimators, and Wiener filtering. There will be exercises to show the use of state-of-the-art DSP simulation software in the solution of signal processing problems.

Practical work reinforces the theoretical concepts introduced.


The course is presented by some of the leading academics in the field of Digital Signal Processing at the University of Surrey.

Dr Ted Chilton

Dr Ted Chilton held the Racal Research Fellowship for four years from 1984-1988 and is now a Senior Lecturer with the Centre for Vision, Speech and Signal Processing (CVSSP). His specialist areas are in speech recognition and the application of transforms to a variety of signal processing problems.

Professor Josef Kittler
Professor Josef Kittler is Director of CVSSP at the University of Surrey. He has more than 30 years of experience in pattern recognition, using both statistical and neural network approaches. He has worked on various theoretical aspects of Pattern Recognition and on many applications including System Identification, Automatic Inspection, ECG diagnosis, Remote Sensing, Robotics, Speech Recognition, Character Recognition, Document Analysis, Biometrics, Image and Video Retrieval and Computer Vision. He has co-authored a book with the title ``Pattern Recognition: a statistical approach'' and published more than 400 papers. He is a member of the Editorial Boards of Pattern Recognition Journal, Image and Vision Computing, Pattern Recognition Letters, Pattern Recognition and Artificial Intelligence, Machine Vision and Applications, and Pattern Analysis and Applications.

Professor Maria Petrou

Professor Maria Petrou received her BSc in Physics from the Aristotle University of Thessaloniki, Greece, and her PhD in Astronomy from the University of Cambridge, UK. She has been working on Image analysis since 1986 and has published more than 250 papers, on Astronomy, Low Level Vision, Feature Extraction, Texture Analysis, Markov Random Fields, Multiresolution optimisation, Probabilistic Relaxation, Colour, Remote Sensing, Industrial Inspection, Medical Signal Processing, etc. She has co-authored a book ``Image Processing: the fundamentals'' published by John Wiley. She is a Chartered Engineer, member of IEEE, Fellow of IEE, Fellow of IAPR. She is Honorary Editor at IEE Electronic Letters and the Treasurer at IAPR.

Dr Philip Jackson

Dr Philip Jackson is lecturer of speech and audio processing in CVSSP, with ten years' experience of signal processing since graduating with honours from Cambridge University Engineering Department. In industry, he developed Active Noise Cancellation systems for commercial aircraft, including the world's first. His PhD, awarded in 2000 by the University of Southampton, was on Acoustic Modelling and Speech Analysis of consonants. He was then research fellow at the University of Birmingham with a project to apply articulatory methods in Automatic Speech Recognition that was rated "outstanding" by EPSRC, before coming to Surrey.


SYLLABUS (provisional- -please bookmark this page)


Day 1 (Dr E Chilton)

Brief overview of the Fourier Transform

The Discrete Fourier transform

The Magnitude and Phase Spectra

The Cosine transform

The Hartley transform

The relationship between the Hartley and Fourier transforms

The Hartley Phase spectrum for phase and time-delay measurement

The Wigner-Ville transform for time-frequency representation

Revision of Signal statistics

The Correlation and Covariance functions

The Autocorrelation Matrix.

Autoregessive and Moving Average Models

Linear Prediction as an Autoregressive model filter

The Yule-Walker Equations

Durbin's Algorithim as a solution to the Yule-Walker equations

The Weiner filter

FIR adaptive filters

The Cost function

The Method of Steepest Descent

Day 2 (Dr P Jackson)


Overview of stochastic time-series approaches

Dynamic time warping

Markov models

Hidden Markov models (HMMs)

- likelihood computation

- optimal path decoding (Viterbi algorithm)

- parameter estimation (Baum-Welch EM algorithm)

Discrete & continuous emission pdfs

Mixtures of Gaussians



Day 3 (Dr E Chilton & Dr P Jackson)


Introduction to Matlab exercises

Laboratory: signal processing in Matlab

Processing and analysis of speech signals

Laboratory: techniques applied to speech


Day 1 (Professor M Petrou)

Higher order statistics and their Applications

Wavelet theory. Wavelet based signal and image processing.

The Uncertainty Principle in signal analysis. Gabor functions

Optimisation techniques: simulated annealing and genetic algorithms.

Day 2 (Professor J Kittler)

Pattern recognition.

Statistical structures and the neuron approach.

Single-layer networks

The Multi-layer Perceptron Neural Network.

Radial based Functions and their applications.

Feature Extraction: classification and Learning.


Provisional Timetable

Part I                

Day 1
Day 2
Day 3
Dr E Chilton
Dr Philip Jackson
Dr Philip Jackson and Dr Edward Chilton
Lecture 1     09:00
Fourier, Cosine and Hartley Transforms

Introduction to Stochastic Models

Matlab Introduction

Coffee             10:30

Lecture 2     11:00
Wigner-Ville trans.
Cepstral analysis

HMMs and Viterbi Algorithm


Lunch             12:30

Lecture 3     13:30
Revision of Signal Statistics, 
Linear prediction

Baum-Welch re-estimation

Introduction to Speech Exercises

Tea                 15:00

Lecture 4    15:30
Wiener filters
FIR adaptive filters
Output Probability Functions


Part II                                      

Day 1 Thursday
Day 2 Friday
Professor Maria Petrou
Professor Josef Kittler
Lecture 1     09:00

Higher order statistics and Applications

Pattern recognition

statistical structures

Coffee             10:30

Lecture 2     11:00

Uncertainty principle applied to signal processing

The neuron approach.

Single-layer networks

Lunch             12:30

Lecture 3     13:30

Wavelet theory.

Wavelet based signal processing

Multi-layer perceptron neural networks
Tea                15:00

Lecture 4    15:50

Optimisation techniques

Simulated annealing genetic algorithms

Radial based functions

Feature extraction

Classification Learning

Revised 20th June 2005



Price per person, including lunch, refreshments and printed course notes
For IEE Members (proof of membership needed)

It is possible to split this course and attend Monday - Wednesday or Thursday - Friday subject to sufficient delegates being registered on the course. Please enquire for price.

Group discounts available for registrations of 3 or more delegates from one company - please telephone Barbara Steel, CE Manager, on 01483 686040 for further information. VAT is not charged as the University is an educational establishment.

Enquiries should be addressed to: Barbara Steel, CE Manager    Tel: +44(0)1483 686040
Fax: +44(0)1483 686041 or send an email by clicking below:
Short Courses Enquiry
To reserve a place on the above course please complete this Registration Form

Short Courses Enquiry

To reserve a place on the above course please complete this Registration Form



Barbara Steel: 15 August 2006