Package rawDeepLearningClassifier.dlClassification.deepAcoustics

Class DeepAcousticsWorker

java.lang.Object

rawDeepLearningClassifier.dlClassification.genericModel.DLModelWorker<StandardPrediction>

rawDeepLearningClassifier.dlClassification.genericModel.GenericModelWorker

rawDeepLearningClassifier.dlClassification.archiveModel.ArchiveModelWorker

rawDeepLearningClassifier.dlClassification.deepAcoustics.DeepAcousticsWorker

public class DeepAcousticsWorker
extends ArchiveModelWorker

The DeepAcousticsWorker class extends ArchiveModelWorker to handle the execution of the Deep Acoustics model. It processes audio data units, runs the model by converting the usual 3D tranformed data into a 4D array (because a colourised image is required), and converts the results into a list of StandardPrediction objects.

Author:

Jamie Macaulay

Field Summary

Fields

Modifier and Type	Field	Description
float	minOverlapThreshold	The minimum overlap threshold for combining bounding boxes.

Fields inherited from class rawDeepLearningClassifier.dlClassification.genericModel.DLModelWorker

MAX_QUEUE_SIZE

Constructor Summary

Constructors

Constructor	Description
DeepAcousticsWorker()

Method Summary

Modifier and Type	Method	Description
float	getMinOverlapThreshold()	Get the minimum overlap threshold for combining bounding boxes.
org.jamdev.jdl4pam.ArchiveModel	loadModel(String currentPath2)	Load a model from a file
ArrayList<StandardPrediction>	runModel(ArrayList<? extends PamDataUnit> dataUnits, float sampleRate, int iChan)	Run the initial data feature extraction and the model
void	setMinOverlapThreshold(float minOverlapThreshold)	Set the minimum overlap threshold for combining bounding boxes.

Methods inherited from class rawDeepLearningClassifier.dlClassification.archiveModel.ArchiveModelWorker

closeModel, getModel, isModelNull, makeModelParams, makeModelResult, prepModel, setModel

Methods inherited from class rawDeepLearningClassifier.dlClassification.genericModel.GenericModelWorker

runModel, setWaveFreqModel

Methods inherited from class rawDeepLearningClassifier.dlClassification.genericModel.DLModelWorker

dataUnits2ModelInput, getModelTransforms, isEnableSoftMax, model2DLTransforms, setEnableSoftMax, setModelTransforms

Methods inherited from class java.lang.Object

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

Field Details

minOverlapThreshold

public float minOverlapThreshold

The minimum overlap threshold for combining bounding boxes.

Constructor Details

DeepAcousticsWorker

public DeepAcousticsWorker()

Method Details

runModel

public ArrayList<StandardPrediction> runModel(ArrayList<? extends PamDataUnit> dataUnits,
 float sampleRate,
 int iChan)

Description copied from class: DLModelWorker

Run the initial data feature extraction and the model

Overrides:

runModel in class DLModelWorker<StandardPrediction>

Parameters:

iChan - - the channel to run the data on.

Returns:

the model to run.

loadModel

public org.jamdev.jdl4pam.ArchiveModel loadModel(String currentPath2)
                                          throws ai.djl.MalformedModelException,
IOException,
ai.djl.engine.EngineException

Description copied from class: ArchiveModelWorker

Load a model from a file

Overrides:

loadModel in class ArchiveModelWorker

Throws:

ai.djl.MalformedModelException

IOException

ai.djl.engine.EngineException

getMinOverlapThreshold

public float getMinOverlapThreshold()

Get the minimum overlap threshold for combining bounding boxes.

Returns:

The minimum overlap threshold.

setMinOverlapThreshold

public void setMinOverlapThreshold(float minOverlapThreshold)

Set the minimum overlap threshold for combining bounding boxes.

Parameters:

minOverlapThreshold - The minimum overlap threshold to set.