org.openimaj.ml.annotation

Class IncrementalAnnotator<OBJECT,ANNOTATION>

java.lang.Object

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

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

Type Parameters:

OBJECT - Type of object

ANNOTATION - Type of annotation

All Implemented Interfaces:

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

Direct Known Subclasses:

AudioAnnotator, FaceRecogniser, FeatureCachingIncrementalBatchAnnotator, InstanceCachingIncrementalBatchAnnotator, KNNAnnotator, NaiveBayesAnnotator

public abstract class IncrementalAnnotator<OBJECT,ANNOTATION>
extends AbstractAnnotator<OBJECT,ANNOTATION>
implements IncrementalTrainer<Annotated<OBJECT,ANNOTATION>>

An Annotator that can be trained/updated incrementally.

Author:

Jonathon Hare (jsh2@ecs.soton.ac.uk)

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	IncrementalAnnotator()

Method Summary

Methods

Modifier and Type	Method and Description
void	train(GroupedDataset<ANNOTATION,? extends ListDataset<OBJECT>,OBJECT> dataset) Train the annotator with the given grouped dataset.
void	train(Iterable<? extends Annotated<OBJECT,ANNOTATION>> data) Train the annotator with the given data.
void	trainMultiClass(GroupedDataset<ANNOTATION,? extends ListDataset<OBJECT>,OBJECT> dataset) Train the annotator with the given grouped dataset.

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

Methods inherited from interface org.openimaj.ml.training.IncrementalTrainer

reset, train

Methods inherited from interface org.openimaj.ml.annotation.Annotator

annotate, getAnnotations

Constructor Detail

IncrementalAnnotator

protected IncrementalAnnotator()

Method Detail

train

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

Train the annotator with the given data. The default implementation of this method just calls IncrementalTrainer.train(Object) on each data item. Subclasses may override to do something more intelligent if necessary.

Specified by:

train in interface IncrementalTrainer<Annotated<OBJECT,ANNOTATION>>

Parameters:

data - the training data

trainMultiClass

public void trainMultiClass(GroupedDataset<ANNOTATION,? extends ListDataset<OBJECT>,OBJECT> dataset)

Train the annotator with the given grouped dataset. This method assumes that each object only appears in a single group of the dataset (i.e. a multi-class problem). Each group corresponds to the one single annotation assigned to each object.

If your dataset contains multiple labels for each object (through an object appearing in multiple groups) you should use train(GroupedDataset).

Parameters:

dataset - the dataset to train on

train

public void train(GroupedDataset<ANNOTATION,? extends ListDataset<OBJECT>,OBJECT> dataset)

Train the annotator with the given grouped dataset. This method assumes that each object can appear in multiple groups of the dataset (i.e. a multi-label problem). Internally, the dataset is converted to a list containing exactly one reference to each object in the dataset with (potentially) multiple annotations.

If the dataset is actually multi-class (i.e. each object belongs to only a single group), then calling this method is equivalent to calling trainMultiClass(GroupedDataset), but is less efficient as the dataset has to be converted into a list.

Some annotator implementations do not care whether the data is multi-class or multi-label, and might choose to override this method to just call trainMultiClass(GroupedDataset) instead.

Parameters:

dataset - the dataset to train on