org.openimaj.ml.annotation.svm

Class SVMAnnotator<OBJECT,ANNOTATION>

java.lang.Object

org.openimaj.ml.annotation.AbstractAnnotator<OBJECT,ANNOTATION>

org.openimaj.ml.annotation.BatchAnnotator<OBJECT,ANNOTATION>

org.openimaj.ml.annotation.svm.SVMAnnotator<OBJECT,ANNOTATION>

Type Parameters:

OBJECT - The object being annotated

ANNOTATION - The type of the annotation

All Implemented Interfaces:

Classifier<ANNOTATION,OBJECT>, Annotator<OBJECT,ANNOTATION>, BatchTrainer<Annotated<OBJECT,ANNOTATION>>

public class SVMAnnotator<OBJECT,ANNOTATION>
extends BatchAnnotator<OBJECT,ANNOTATION>

Wraps the libsvm SVM and provides basic positive/negative annotation for a single class.

Author:

David Dupplaw (dpd@ecs.soton.ac.uk)

Field Summary

Fields

Modifier and Type	Field and Description
HashMap<Integer,ANNOTATION>	classMap Stores the mapping between the positive and negative class and the annotation
static int	NEGATIVE_CLASS The input to the SVM model for negative classes
static int	POSITIVE_CLASS The input to the SVM model for positive classes

Constructor Summary

Constructors

Constructor and Description
SVMAnnotator(FeatureExtractor<? extends FeatureVector,OBJECT> extractor) Constructor that takes the feature extractor to use.

Method Summary

Methods

Modifier and Type	Method and Description
List<ScoredAnnotation<ANNOTATION>>	annotate(OBJECT object) Generate annotations for the given object.
double	crossValidation(List<? extends Annotated<OBJECT,ANNOTATION>> data, int numFold) Performs cross-validation on the SVM.
static double	crossValidation(libsvm.svm_problem prob, libsvm.svm_parameter param, int numFold) Performs cross-validation on the SVM.
Set<ANNOTATION>	getAnnotations()
void	loadModel(File loadModel) Load an existing svm model.
void	setSaveModel(File saveModel) Set whether to save the SVM model to disk.
void	train(List<? extends Annotated<OBJECT,ANNOTATION>> data) Train the object with the given data.

Methods inherited from class org.openimaj.ml.annotation.BatchAnnotator

train

Methods inherited from class org.openimaj.ml.annotation.AbstractAnnotator

classify

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

POSITIVE_CLASS

public static final int POSITIVE_CLASS

The input to the SVM model for positive classes

See Also:

Constant Field Values

NEGATIVE_CLASS

public static final int NEGATIVE_CLASS

The input to the SVM model for negative classes

See Also:

Constant Field Values

classMap

public HashMap<Integer,ANNOTATION> classMap

Stores the mapping between the positive and negative class and the annotation

Constructor Detail

SVMAnnotator

public SVMAnnotator(FeatureExtractor<? extends FeatureVector,OBJECT> extractor)

Constructor that takes the feature extractor to use.

Parameters:

extractor - The feature extractor

Method Detail

train

public void train(List<? extends Annotated<OBJECT,ANNOTATION>> data)

Train the object with the given data. BatchTrainers should reset the underlying model everytime this method is called.

Parameters:

data - the training data

See Also:

BatchTrainer.train(java.util.List)

getAnnotations

public Set<ANNOTATION> getAnnotations()

Returns:

a Set of all annotations this Annotator knows about

See Also:

Annotator.getAnnotations()

annotate

public List<ScoredAnnotation<ANNOTATION>> annotate(OBJECT object)

Generate annotations for the given object.

Parameters:

object - the image

Returns:

generated annotations

See Also:

Annotator.annotate(java.lang.Object)

setSaveModel

public void setSaveModel(File saveModel)

Set whether to save the SVM model to disk.

Parameters:

saveModel - The file name to save to, or null to disable saving.

loadModel

public void loadModel(File loadModel)
               throws IOException

Load an existing svm model.

Parameters:

loadModel - The model to load from

Throws:

IOException - If the loading does not complete

crossValidation

public double crossValidation(List<? extends Annotated<OBJECT,ANNOTATION>> data,
                     int numFold)

Performs cross-validation on the SVM.

Parameters:

data - The data

numFold - The number of folds

Returns:

The calculated accuracy

crossValidation

public static double crossValidation(libsvm.svm_problem prob,
                     libsvm.svm_parameter param,
                     int numFold)

Performs cross-validation on the SVM.

Parameters:

prob - The problem

param - The parameters

numFold - The number of folds

Returns:

The accuracy