net.semanticmetadata.lire.utils

类 MetricsUtils

java.lang.Object

net.semanticmetadata.lire.utils.MetricsUtils

public class MetricsUtils
extends java.lang.Object

User: mlux Date: 25.11.2009 Time: 14:32:49

构造器概要

构造器

构造器和说明
MetricsUtils()

方法概要

限定符和类型	方法和说明
static double	chisquare(double[] d1, double[] d2) Chi^2 statistics.
static double	cosineCoefficient(double[] hist1, double[] hist2)
static double	distL1(byte[] h1, byte[] h2)
static double	distL1(double[] h1, double[] h2)
static double	distL1(float[] h1, float[] h2)
static double	distL1(int[] h1, int[] h2) Manhattan distance
static double	distL2(double[] h1, double[] h2) Euclidean distance
static double	distL2(float[] h1, float[] h2) Euclidean distance
static double	distL2(int[] h1, int[] h2) Euclidean distance
static double	jsd(byte[] h1, byte[] h2)
static double	jsd(double[] h1, double[] h2)
static double	jsd(float[] h1, float[] h2)
static double	jsd(int[] h1, int[] h2) Jeffrey Divergence or Jensen-Shannon divergence (JSD) from Deselaers, T.; Keysers, D. & Ney, H.
static double	ksDistance(double[] d1, double[] d2) Kolmogorov-Smirnoff Distance for two equal length, equal summed histograms as described in Rubner, Yossi, Carlo Tomasi, and Leonidas J.
static double[]	normalizeL1(double[] histogram) Euclidean normalization of a double[] histogram. // todo: make it faster and less memory consuming ...
static double[]	normalizeL2(double[] histogram) Euclidean normalization of a double[] histogram. // todo: make it faster and less memory consuming ...
static double[]	normalizeMax(double[] histogram) Max normalization of a double[] histogram. // todo: make it faster and less memory consuming ...
static double	simpleEMD(double[] d1, double[] d2) Earth Mover's Distance for two equal length, equal summed histograms as described in Rubner, Yossi, Carlo Tomasi, and Leonidas J.
static double	tanimoto(double[] h1, double[] h2)
static double	tanimoto(float[] h1, float[] h2)
static double	tanimoto(int[] h1, int[] h2)

从类继承的方法 java.lang.Object

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

构造器详细资料

MetricsUtils

public MetricsUtils()

方法详细资料

distL1

public static double distL1(int[] h1,
                            int[] h2)

Manhattan distance

参数:

h1 -

h2 -

返回:

distL1

public static double distL1(double[] h1,
                            double[] h2)

distL2

public static double distL2(int[] h1,
                            int[] h2)

Euclidean distance

参数:

h1 -

h2 -

返回:

distL2

public static double distL2(double[] h1,
                            double[] h2)

Euclidean distance

参数:

h1 -

h2 -

返回:

distL2

public static double distL2(float[] h1,
                            float[] h2)

Euclidean distance

参数:

h1 -

h2 -

返回:

jsd

public static double jsd(int[] h1,
                         int[] h2)

Jeffrey Divergence or Jensen-Shannon divergence (JSD) from Deselaers, T.; Keysers, D. & Ney, H. Features for image retrieval: an experimental comparison Inf. Retr., Kluwer Academic Publishers, 2008, 11, 77-107

参数:

h1 -

h2 -

返回:

chisquare

public static double chisquare(double[] d1,
                               double[] d2)

Chi^2 statistics.

参数:

d1 -

d2 -

返回:

distance like "unlikelihood"

simpleEMD

public static double simpleEMD(double[] d1,
                               double[] d2)

Earth Mover's Distance for two equal length, equal summed histograms as described in Rubner, Yossi, Carlo Tomasi, and Leonidas J. Guibas. "The earth mover's distance as a metric for image retrieval." International journal of computer vision 40.2 (2000): 99-121.

参数:

d1 - NOTE: sum(d1) needs to be equal to sum(d2)

d2 -

返回:

EMD

ksDistance

public static double ksDistance(double[] d1,
                                double[] d2)

Kolmogorov-Smirnoff Distance for two equal length, equal summed histograms as described in Rubner, Yossi, Carlo Tomasi, and Leonidas J. Guibas. "The earth mover's distance as a metric for image retrieval." International journal of computer vision 40.2 (2000): 99-121.

参数:

d1 - NOTE: sum(d1) needs to be equal to sum(d2)

d2 -

返回:

EMD

jsd

public static double jsd(byte[] h1,
                         byte[] h2)

jsd

public static double jsd(float[] h1,
                         float[] h2)

jsd

public static double jsd(double[] h1,
                         double[] h2)

tanimoto

public static double tanimoto(int[] h1,
                              int[] h2)

tanimoto

public static double tanimoto(float[] h1,
                              float[] h2)

tanimoto

public static double tanimoto(double[] h1,
                              double[] h2)

cosineCoefficient

public static double cosineCoefficient(double[] hist1,
                                       double[] hist2)

distL1

public static double distL1(float[] h1,
                            float[] h2)

distL1

public static double distL1(byte[] h1,
                            byte[] h2)

normalizeMax

public static double[] normalizeMax(double[] histogram)

Max normalization of a double[] histogram. // todo: make it faster and less memory consuming ...

参数:

histogram -

返回:

normalizeL2

public static double[] normalizeL2(double[] histogram)

Euclidean normalization of a double[] histogram. // todo: make it faster and less memory consuming ...

参数:

histogram -

返回:

normalizeL1

public static double[] normalizeL1(double[] histogram)

Euclidean normalization of a double[] histogram. // todo: make it faster and less memory consuming ...

参数:

histogram -

返回: