ARTIFICIAL NEURAL NETWORKS AND PATTERN RECOGNITION

Basics of Pattern Recognition

The pattern is an object, process or event Patterns are made up of features, which are measurements used as inputs to the classification system. A class is a set of patterns that share common attribute(features) usually from the same information source During Recognition(or classification) classes are assigned to the objects A classifier is a machine that performs such task

What is a Pattern?

A pattern is the opposite of chaos; it is an entity vaguely defined ,that could be given a name

Finger Prints

Music Pattern
Pattern of Constellations

Other Patterns..
Speech Recognition Weather Patterns Financial Series Patterns Medical Diagnostics-ECG ,EEG Face Recognition Optical Character-Eg. Reading machines for blind people Natural Language

Thus ,it can be said that Pattern recognition is the study of how machines can observe the environment ,learn to distinguish between the patterns of interest and finally make sound and reasonable decisions about the categories of patterns.

An Example

Problem: Sorting incoming fish on a conveyor belt according to species Assume that we have only two kinds of sh: sea bass salmon
Picture of two species of fish

Decision Process

What kind of information can distinguish one species from the other? length, width, weight, number and shape of fins, tail shape, etc. What can cause problems during sensing? lighting conditions, position of sh on the conveyor belt, camera noise, dirt on background etc. What are the steps in the process? capture image isolate sh take measurements make decision and extract relevant data

PreProcessing

Feature Extraction

Salmon Sea Baas

Classification

Preprocessin g
Segment the sh from Background

Feature Extraction
Image data from each sh Summarized by feature extractor whose purpose isto reduce the data by measuring

Classication
The features are passed to a classier, that usesthe featuresto decide which classthe instance belongs to.

Selecting Features

Assume a fisherman told us that a sea bass is generally longer than a salmon. We can use length as a feature and decide between sea bass and salmon according to a threshold on length. These feature measurements are passed to the classifier. But how can we choose this threshold?

Training data: Class+features

Selecting Length is a poor feature.

Selecting lightness of the scales may prove helpful which reduces the error rate

Cost of Error

We should also consider costs of different errors we make in our decisions. For example, if the sh packing company knows that: Customers who buy salmon will object vigorously if they see sea bass in their cans. Customers who buy sea bass will not be unhappy if they occasionally see some expensive salmon in their cans. How does this knowledge aect our decision?

Multiple Features
We can use two features in our decision: 1. lightness: x1 2. width: x2 Each sh image is now represented as a point (feature vector) x= x1 in 2D feature space

x2

Scatter Plot Highlightin g the decision boundary for 2 classes of fish

Process Diagram of Pattern Recognition System

Converting an Image

Camera captures an image Image needs to be converted to a form that can be processed by the Neural Network Image consists of pixels Values can be assigned to color of each pixel A vector can represent the pixel values in an image

Assign 1 if pixel is filled, otherwise assign 0

Application of Pattern Recognition

Signature Verification and Character Recognition Signature is an identification of a person

through his/her hand writing. It also plays an important role in many applications concerned with security, access control or financial and contractual matters. Optical Character Recognition Input: scanned images of printed text Output: Computer readable version of input contents

Application of Pattern Recognition

Speech Recognition

Signal processing:

Convert the audio wave into a sequence of feature vectors

Speech recognition:
Decode the sequence of feature vectors into a sequence of words
Determine the meaning of the recognized words

Semantic interpretation:

Dialog Management:

Correct errors and help get the task done

What words to use to maximize user understanding Generate synthetic speech from a marked-up word string

Response Generation Speech synthesis (Text to Speech):

Overview

Applications related to Speech Recognition

Speech Recognition Figure out what a person is saying. Speaker Verification Authenticate that a person is who she/he claims to be. Limited speech patterns Speaker Identification Assigns an identity to the voice of an unknown person. Arbitrary speech patterns

Thank You..