DHSCH 1

|

|
by R. O. Duda, P.
E. Hart and D. G. Stork, John Wiley & Sons, 2000

hapter 1: Introduction to Pattern
Recognition (Sections 1.1-1.6)
Machine Perception
An Example
Pattern Recognition Systems
The Design ycle
Learning and Adaptation
onclusion

Machine Perception
Build a machine that can recognize patterns:

Speech recognition
Fingerprint identification
OR (Optical haracter Recognition)
DNA sequence identification
Pattern lassification, hapter 1

An Example
³Sorting incoming Fish on a conveyor according to

species using optical sensing´
Sea bass
Species
Salmon

e
Problem Analysis
Set up a camera and take some sample images to extract
features
Length
Lightness
Width
Number and shape of fins
Position of the mouth, etc«
This is the set of all suggested features to explore for use in our
classifier!

õ
Preprocessing
Use a segmentation operation to isolate fishes from one

another and from the background
Information from a single fish is sent to a feature

extractor whose purpose is to reduce the data by
measuring certain features
The features are passed to a classifier

Á

Ø
lassification
Select the length of the fish as a possible feature for
discrimination

ï

The length is a poor feature alone!
Select the lightness as a possible feature.


Threshold decision boundary and cost relationship

Move our decision boundary toward smaller values of
lightness in order to minimize the cost (reduce the number
of sea bass that are classified salmon!)
Task of decision theory

Adopt the lightness and add the width of the fish

Fish : D::
h


e
We might add other features that are not correlated

with the ones we already have. A precaution should
be taken not to reduce the performance by adding
such ³noisy features´
Ideally, the best decision boundary should be the one

which provides an optimal performance such as in the
following figure:

õ

Á
However, our satisfaction is premature because

the central aim of designing a classifier is to
correctly classify novel input
Issue of generalization!

Ø

ï
Pattern Recognition Systems
Sensing
Use of a transducer (camera or microphone)
PR system depends of the bandwidth, the resolution
sensitivity distortion of the transducer
Segmentation and grouping

Patterns should be well separated and should not overlap


Feature extraction
Discriminative features
Invariant features with respect to translation, rotation and
scale.
lassification
Use a feature vector provided by a feature extractor to
assign the object to a category
Post Processing
Exploit context input dependent information other than from
the target pattern itself to improve performance

The Design ycle
Data collection
Feature hoice
Model hoice
Training
Evaluation
omputational omplexity


Data ollection
How do we know when we have collected an adequately
large and representative set of examples for training and
testing the system?

e
Feature hoice
Depends on the characteristics of the problem domain.
Simple to extract, invariant to irrelevant transformation
insensitive to noise.

õ
Model hoice
Unsatisfied with the performance of our fish classifier and
want to jump to another class of model

Á
Training
Use data to determine the classifier. Many different
procedures for training classifiers and choosing models

Ø
Evaluation
Measure the error rate (or performance and switch from
one set of features to another one

ï
omputational omplexity
What is the trade-off between computational ease and
performance?
(How an algorithm scales as a function of the number of

features, patterns or categories?)

Learning and Adaptation
Supervised learning
A teacher provides a category label or cost for each
pattern in the training set
Unsupervised learning
The system forms clusters or ³natural groupings´ of the
input patterns

onclusion
Reader seems to be overwhelmed by the number,

complexity and magnitude of the sub-problems of
Pattern Recognition
Many of these sub-problems can indeed be solved

Many fascinating unsolved problems still remain

DHSCH 1

Uploaded by

Copyright:

Available Formats

DHSCH 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

DHSCH 1

Uploaded by

Copyright:

Available Formats

| 

    

 Build a machine that can recognize patterns:

Pattern lassification, hapter 1

 ³Sorting incoming Fish on a conveyor according to

Pattern lassification, hapter 1

Pattern lassification, hapter 1

 Use a segmentation operation to isolate fishes from one

 Information from a single fish is sent to a feature

 The features are passed to a classifier

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

The length is a poor feature alone!

Select the lightness as a possible feature.

Pattern lassification, hapter 1

Pattern lassification, hapter 1

 Threshold decision boundary and cost relationship

Task of decision theory

Pattern lassification, hapter 1

 Adopt the lightness and add the width of the fish

Pattern lassification, hapter 1

Pattern lassification, hapter 1

 We might add other features that are not correlated

 Ideally, the best decision boundary should be the one

Pattern lassification, hapter 1

Pattern lassification, hapter 1

 However, our satisfaction is premature because

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern Recognition Systems

 Segmentation and grouping

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

The Design ycle

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

 (How an algorithm scales as a function of the number of

Pattern lassification, hapter 1

Learning and Adaptation

Pattern lassification, hapter 1

 Reader seems to be overwhelmed by the number,

 Many of these sub-problems can indeed be solved

Pattern lassification, hapter 1

You might also like

|

Build a machine that can recognize patterns:

Pattern lassification, hapter 1

³Sorting incoming Fish on a conveyor according to

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Use a segmentation operation to isolate fishes from one

Information from a single fish is sent to a feature

The features are passed to a classifier

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Threshold decision boundary and cost relationship

Pattern lassification, hapter 1

Adopt the lightness and add the width of the fish

Pattern lassification, hapter 1

Pattern lassification, hapter 1

We might add other features that are not correlated

Ideally, the best decision boundary should be the one

Pattern lassification, hapter 1

Pattern lassification, hapter 1

However, our satisfaction is premature because

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Segmentation and grouping

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

The Design ycle

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Pattern lassification, hapter 1

(How an algorithm scales as a function of the number of

Pattern lassification, hapter 1

Pattern lassification, hapter 1

Reader seems to be overwhelmed by the number,

Many of these sub-problems can indeed be solved

Pattern lassification, hapter 1