Package matchedTemplateClassifer

Class MTClassifier

java.lang.Object

matchedTemplateClassifer.MTClassifier

All Implemented Interfaces:

Serializable, Cloneable, ManagedParameters

Direct Known Subclasses:

DefaultTemplates

public class MTClassifier
extends Object
implements Serializable, Cloneable, ManagedParameters

Parameters and useful functions for a single MT classifier.

Author:

Jamie Macaulay

See Also:

• Serialized Form

Field Summary

Fields

Modifier and Type	Field	Description
double	currentSr	The current sample rates of the waveforms
int	normalisation	The normalisation
static final int	TEST_FFT_LENGTH
double	thresholdToAccept	Threshold to classify
MatchTemplate	waveformMatch	The waveform to match
MatchTemplate	waveformReject	The waveform to reject

Constructor Summary

Constructors

Constructor	Description
MTClassifier()	Default MT classifier

Method Summary

Modifier and Type	Method	Description
double[]	binArray(int start, int stop, double step)
MatchedTemplateResult	calcCorrelationMatch(ComplexArray click, float sR)	Calculate a value for the click reject and match template ration correlation
us.hebi.matlab.mat.types.Struct	calcCorrelationMatchTest(ComplexArray click, float sR)	Tests the correlation function by outputting data to .mat file.
static double[]	conditionClickLength(double[] click, int fftLength, int smoothing)	The click needs to be the same or shorter than the FFT length.
int	getFFTLength(float sR)	Deprecated.
PamParameterSet	getParameterSet()	Get the auto-generated parameter set, and then add in the fields that are not included because they are not public and do not have getters.
ComplexArray	getWaveformMatchFFT(float sR, int length)	Get the match waveform FFT for the sampleRate.
ComplexArray	getWaveformRejectFFT(float sR, int length)	Get the match waveform for the sample rate
static double[]	normaliseWaveform(double[] waveform, int normeType)	Normalise the waveform.
void	reset()	Reset all interpolated waveforms etc.

Methods inherited from class java.lang.Object

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

Field Details

thresholdToAccept

public double thresholdToAccept

Threshold to classify

normalisation

public int normalisation

The normalisation

currentSr

public double currentSr

The current sample rates of the waveforms

waveformMatch

public MatchTemplate waveformMatch

The waveform to match

waveformReject

public MatchTemplate waveformReject

The waveform to reject

TEST_FFT_LENGTH

public static final int TEST_FFT_LENGTH

See Also:

• Constant Field Values

Constructor Details

MTClassifier

public MTClassifier()

Default MT classifier

Method Details

getWaveformMatchFFT

public ComplexArray getWaveformMatchFFT(float sR,
 int length)

Get the match waveform FFT for the sampleRate.

Parameters:

sR - - the sample rate in samples per second

fftLength - - the FFTlength to use. Bins around waveform peak are used.

getWaveformRejectFFT

public ComplexArray getWaveformRejectFFT(float sR,
 int length)

Get the match waveform for the sample rate

Parameters:

sR - - the sample rate in samples per second

normaliseWaveform

public static double[] normaliseWaveform(double[] waveform,
 int normeType)

Normalise the waveform.

Parameters:

waveform -

normeType -

Returns:

getFFTLength

@Deprecated
public int getFFTLength(float sR)

Deprecated.

Get the FFT length which should be used for the current sample rate.

Parameters:

length -

the - sample rate.

Returns:

the FFT length to be used.This is the maximum length of the two templates.

reset

public void reset()

Reset all interpolated waveforms etc.

calcCorrelationMatchTest

public us.hebi.matlab.mat.types.Struct calcCorrelationMatchTest(ComplexArray click,
 float sR)

Tests the correlation function by outputting data to .mat file. Calculate a value for the click reject and match template ration correlation.

Parameters:

click - - the input click to perform correlation match on. It should be the same length and sample rate as the templates.

sR - - the sample rate in samples per second

Returns:

an ML Structure with results of all stuff.

calcCorrelationMatch

public MatchedTemplateResult calcCorrelationMatch(ComplexArray click,
 float sR)

Calculate a value for the click reject and match template ration correlation

Parameters:

click - - the input click FFT to perform correlation match on. It should be the same length and sample rate as the templates.

sR - - the sample rate in samples per second

Returns:

the correlation reject match template.

binArray

public double[] binArray(int start,
 int stop,
 double step)

conditionClickLength

public static double[] conditionClickLength(double[] click,
 int fftLength,
 int smoothing)

The click needs to be the same or shorter than the FFT length. If the click is longer need to find the click in the waveform snipper and extract a subsection.

getParameterSet

public PamParameterSet getParameterSet()

Get the auto-generated parameter set, and then add in the fields that are not included because they are not public and do not have getters. Note: for each field, we search the current class and (if that fails) the superclass. It's done this way because MTClassifier might be used directly (and thus the field would be found in the class) and it also might be used as a superclass to something else (e.g. DefaultTemplates) in which case the field would only be found in the superclass.

Specified by:

getParameterSet in interface ManagedParameters

Returns:

description of the parameters in a class.