1996
@phdthesis{Squ1996,
vgclass = {thesis},
vgproject = {nn,invariance},
author = {David McG. Squire},
title = {Model-based Neural Networks for Invariant Pattern
Recognition},
school = {School of Computing, Curtin University of Technology},
address = {Perth, Western Australia},
month = {October},
year = {1996},
note = {(Submitted 1996, awarded March 19th, 1997)},
url = {/publications/thesis.html},
abstract = {In this thesis, the notion of \emph{Model-Based Neural
Networks} is introduced. Model-Based Neural Networks, whilst retaining
the essential structure and advantages of traditional neural networks,
enable explicit modeling of the target system. This can result in
dramatically improved generalization of classification performance to
patterns not present in the training data, and representation in
considerably lower-dimensional state spaces.
An important problem in many areas of computer vision is invariant
pattern recognition. This is the chosen domain for demonstrating the
efficacy of the Model-Based Neural Network approach. Model-Based Neural
Networks can be constructed which have responses which are invariant to
specified transformations of the input data. Such networks can be
trained with much smaller training sets than are required by
traditional networks, since it is no longer necessary to provide
examples of transformed versions of the input prototypes. This, coupled
with the reduction in the dimensionality of the parameter space, means
that training such networks is often much less
computationally-expensive than the traditional alternative.
To situate this work, a review of existing general techniques for
invariant pattern recognition is presented in Chapter~2, and of
previous neural network-based approaches in Chapter~3. Chapter~4
presents a variety of different forms of Model-Based Neural Network,
and demonstrates their utility on a range of invariant pattern
recognition problems, both real and synthetic. Included is a comparison
with an earlier study, which reveals the great improvements possible
with Model-Based Neural Networks.
Chapter~5 introduces a new invariant feature of two-dimensional
contours, the Invariance Signature. It is shown in Chapter~6 that a
Model-Based Neural Network can be constructed which calculates this
multi-dimensional feature, and classifies patterns on this basis.
Chapter~7 reports experimental results demonstrating that such
Invariance Signature-based MBNNs can be employed successfully for
shift-, rotation- and scale-invariant optical character recognition.},
}