/**
* Tun ester2 diarization.
*
* @param parameter the parameter
* @param clusterSet the cluster set
* @return the tree map
* @throws DiarizationException the diarization exception
* @throws Exception the exception
*/
public TreeMap<String, DiarizationResultList> tunEster2Diarization(
Parameter parameter, ClusterSet clusterSet) throws DiarizationException, Exception {
TreeMap<String, DiarizationResultList> result = new TreeMap<String, DiarizationResultList>();
// double paramThr = parameter.getParameterClustering().getThreshold();
lMin = parameter.getParameterDiarization().getThreshold("l");
lMax = parameter.getParameterDiarization().getMaxThreshold("l");
hMin = parameter.getParameterDiarization().getThreshold("h");
hMax = parameter.getParameterDiarization().getMaxThreshold("h");
dMin = parameter.getParameterDiarization().getThreshold("d");
dMax = parameter.getParameterDiarization().getMaxThreshold("d");
cMin = parameter.getParameterDiarization().getThreshold("c");
cMax = parameter.getParameterDiarization().getMaxThreshold("c");
String featureDesc = parameter.getParameterInputFeature().getFeaturesDescriptorAsString();
AudioFeatureSet featureSet = null;
ClusterSet clustersSegInit = null;
if (parameter.getParameterDiarization().isLoadInputSegmentation() == false) {
featureSet = loadFeature(parameter, clusterSet, featureDesc);
featureSet.setCurrentShow(parameter.show);
int nbFeatures = featureSet.getNumberOfFeatures();
clusterSet.getFirstCluster().firstSegment().setLength(nbFeatures);
clustersSegInit = sanityCheck(clusterSet, featureSet, parameter);
} else {
featureSet = loadFeature(parameter, clusterSet, featureDesc);
featureSet.setCurrentShow(parameter.show);
clustersSegInit = sanityCheck(clusterSet, featureSet, parameter);
featureSet = loadFeature(parameter, clustersSegInit, featureDesc);
featureSet.setCurrentShow(parameter.show);
}
// seg IRIT
// ClusterSet clustersSegSave = clustersSegInit;
// seg IRIT
ClusterSet referenceClusterSet = MainTools.readTheSecondClusterSet(parameter);
ClusterSet uemClusterSet = MainTools.readThe3rdClusterSet(parameter);
if (parameter.getParameterDiarization().isLastStepOnly()) {
String key = "l=" + lMin + " h=" + hMin + " d=" + dMin;
DiarizationResultList values = null;
if (parameter.getParameterDiarization().isCEClustering() == false) {
logger.warning(" nothing to do isCEClustering == false");
} else {
values =
tunEster2SpeakerCLRClustering(
referenceClusterSet,
uemClusterSet,
key,
"ce",
clusterSet,
clusterSet,
featureSet,
parameter);
}
result.put(key, values);
return result;
}
ClusterSet clustersSegSave =
segmentation("GLR", "FULL", clustersSegInit, featureSet, parameter);
for (double l = lMin; l <= lMax; l += 0.5) {
ClusterSet clustersSeg = clustersSegSave.clone();
logger.finest("clustering l=" + l);
ClusterSet clustersLClust = clusteringLinear(l, clustersSeg, featureSet, parameter);
// ---- Begin NEW v 1.14 ---
for (double h = hMin; h <= hMax; h += 0.5) {
// for (double h = hMin; h <= hMax; h += 0.2) {
// ---- end NEW v 1.14 ---
// if (h > l) {
ClusterSet clustersHClust = clustering(h, clustersLClust, featureSet, parameter);
for (double d = dMin; d <= dMax; d += 50) {
ClusterSet clustersDClust = decode(8, d, clustersHClust, featureSet, parameter);
// double error = DiarizationError.scoreOfMatchedSpeakers(referenceClusterSet,
// clustersDClust);
ClusterSet clustersSplitClust =
speech(
"10,10,50", clusterSet, clustersSegInit, clustersDClust, featureSet, parameter);
ClusterSet clustersGender = gender(clusterSet, clustersSplitClust, featureSet, parameter);
String key = "l=" + l + " h=" + h + " d=" + d;
DiarizationResultList values = null;
if (parameter.getParameterDiarization().isCEClustering() == false) {
values = new DiarizationResultList(0, 0, 1);
DiarizationError computeError =
new DiarizationError(referenceClusterSet, uemClusterSet);
DiarizationResult error = computeError.scoreOfMatchedSpeakers(clustersGender);
values.setResult(0, 0, error);
logger.finer(parameter.show + " key=" + key + " resultat du fichier");
values.log("partial result: " + parameter.show + " " + key);
} else {
// V4.19 = CLUST_H_BIC_GMM_MAP
// values = tunEster2SpeakerCLRClustering(referenceClusterSet, key, "bicgmmmap",
// clustersGender, clustersGender, featureSet, parameter);
// V5.16 = ce_d
// values = tunEster2SpeakerCLRClustering(referenceClusterSet, key, "ce_d",
// clustersGender, clustersGender, featureSet, parameter);
values =
tunEster2SpeakerCLRClustering(
referenceClusterSet,
uemClusterSet,
key,
"ce",
clustersGender,
clustersGender,
featureSet,
parameter);
}
if (result.containsKey(key)) {
result.get(key).addResultArray(values);
} else {
result.put(key, values);
}
}
// }
}
}
return result;
}
/**
* Tun ester2 speaker clr clustering.
*
* @param referenceClusterSet the reference cluster set
* @param uemClusterSet the uem cluster set
* @param partialKey the partial key
* @param method the method
* @param clusterSetBase the cluster set base
* @param clusterSet the cluster set
* @param featureSet the feature set
* @param parameter the parameter
* @return the diarization result list
* @throws Exception the exception
*/
public DiarizationResultList tunEster2SpeakerCLRClustering(
ClusterSet referenceClusterSet,
ClusterSet uemClusterSet,
String partialKey,
String method,
ClusterSet clusterSetBase,
ClusterSet clusterSet,
AudioFeatureSet featureSet,
Parameter parameter)
throws Exception {
String oldSpeechDetectorMethod = parameter.getParameterInputFeature().getSpeechMethodAsString();
double oldSpeechDetectorThreshold = parameter.getParameterInputFeature().getSpeechThreshold();
String oldModelKind = parameter.getParameterModel().getModelKindAsString();
int oldNumberOfComponent = parameter.getParameterModel().getNumberOfComponents();
String oldMethod = parameter.getParameterClustering().getMethodAsString();
double oldThreshold = parameter.getParameterClustering().getThreshold();
String oldEMControl = parameter.getParameterEM().getEMControl();
int oldNTop = parameter.getParameterTopGaussian().getScoreNTop();
boolean oldSaveAll = parameter.getParameterDiarization().isSaveAllStep();
DiarizationResultList localResult = new DiarizationResultList(cMin, cMax, mult);
DiarizationError computeError = new DiarizationError(referenceClusterSet, uemClusterSet);
double prevScore = cMin;
// ---- Begin NEW v 1.13 ---
parameter.getParameterInputFeature().setSpeechMethod("E");
parameter.getParameterInputFeature().setSpeechThreshold(0.1);
// ---- End NEW v 1.13 ---
String FeatureFormat = "featureSetTransformation";
String dir = "ester2";
InputStream ubmInputStream = getClass().getResourceAsStream(dir + "/ubm.gmm");
GMMArrayList ubmVect =
MainTools.readGMMContainer(ubmInputStream, parameter.getParameterModel());
GMM ubm = ubmVect.get(0);
/*
* int nbCep = 16; logger.info("---> nbCep:"+nbCep); FeatureSet featureSet2 = loadFeature(parameter, clusterSet, "audio16kHz2sphinx,1:3:2:0:0:0,"+nbCep+",1:1:300:4"); logger.info("---> nbCep:"+nbCep); //logger.fine("*** nbFeaturesNorm:" +
* clusterSetBase.getLength());
*/
// A tester
AudioFeatureSet featureSet2 =
loadFeature(featureSet, parameter, clusterSet, FeatureFormat + ",1:3:2:0:0:0,13,1:1:300:4");
// v5.14
// FeatureSet featureSet2 = loadFeature(featureSet, parameter, clusterSet, FeatureFormat
// + ",1:3:2:0:0:0,13,1:1:0:0");
parameter.getParameterModel().setModelKind("DIAG");
parameter.getParameterModel().setNumberOfComponents(ubm.getNbOfComponents());
parameter.getParameterClustering().setMethod(method);
// ---- Begin NEW v 1.19 ---
// parameter.getParameterEM().setEMControl("1,1,0.01");
// parameter.getParameterClustering().setThreshold(0);
// ---- End NEW v 1.19 ---
parameter.getParameterClustering().setThreshold(cMax);
parameter.getParameterEM().setEMControl("1,5,0.01");
parameter.getParameterTopGaussian().setScoreNTop(5);
parameter.getParameterDiarization().setSaveAllStep(false);
CLRHClustering clustering = new CLRHClustering(clusterSet, featureSet2, parameter, ubm);
// int nbCluster = clusterSet.clusterGetSize();
// logger.info("initialise clustering CLR clusterSet:"+clusterSet);
clustering.initialize();
double score = clustering.getScoreOfCandidatesForMerging();
DiarizationResult error = computeError.scoreOfMatchedSpeakers(clustering.getClusterSet());
double errorRate = error.getErrorRate();
localResult.setResult(prevScore, score, error);
// prevScore = Math.max(score, prevScore);
logger.fine(
"first "
+ parameter.show
+ " key="
+ partialKey
+ " clrScore="
+ score
+ " clrErrorRate="
+ errorRate
+ " clrSize="
+ clustering.getSize()
+ "/"
+ referenceClusterSet.clusterGetSize());
while ((score < cMax) && (clustering.getSize() > 1)) {
localResult.setResult(prevScore, score, error);
prevScore = Math.max(score, prevScore);
clustering.mergeCandidates();
// -- start V5.16 --
// logger.info("--> Decoding");
// ClusterSet decodeClusterSet = MDecode.make(featureSet2, clustering.getClusterSet(),
// clustering.getGmmList(), parameter);
// logger.info("--> Clustering");
// featureSet2 = loadFeature(featureSet, parameter, decodeClusterSet, FeatureFormat
// + ",1:3:2:0:0:0,13,1:1:300:4");
// clustering = new CLRHClustering(decodeClusterSet, featureSet2, parameter, ubm);
// clustering.initialize();
// -- end V5.16 --
score = clustering.getScoreOfCandidatesForMerging();
error = computeError.scoreOfMatchedSpeakers(clustering.getClusterSet());
errorRate = error.getErrorRate();
// localResult.setResult(prevScore, score, error);
// prevScore = Math.max(score, prevScore);
logger.fine(
parameter.show
+ " key="
+ partialKey
+ " clrScore="
+ score
+ " clrErrorRate="
+ errorRate
+ " clrSize="
+ clustering.getSize()
+ "/"
+ referenceClusterSet.clusterGetSize());
}
localResult.setResult(prevScore, score, error);
localResult.setResult(score, cMax, error);
logger.finer(parameter.show + " key=" + partialKey + " resultat du fichier");
localResult.log("partial result: " + parameter.show + " " + partialKey);
clustering.reset();
parameter.getParameterModel().setNumberOfComponents(oldNumberOfComponent);
parameter.getParameterModel().setModelKind(oldModelKind);
parameter.getParameterClustering().setMethod(oldMethod);
parameter.getParameterClustering().setThreshold(oldThreshold);
parameter.getParameterEM().setEMControl(oldEMControl);
parameter.getParameterTopGaussian().setScoreNTop(oldNTop);
parameter.getParameterInputFeature().setSpeechMethod(oldSpeechDetectorMethod);
parameter.getParameterInputFeature().setSpeechThreshold(oldSpeechDetectorThreshold);
parameter.getParameterDiarization().setSaveAllStep(oldSaveAll);
return localResult;
}
/**
* Ester2 version.
*
* @param parameter the parameter
* @throws DiarizationException the diarization exception
* @throws Exception the exception
*/
public void ester2Version(Parameter parameter) throws DiarizationException, Exception {
// ** Caution this system is developed using Sphinx MFCC computed with legacy mode
ClusterSet referenceClusterSet = null;
if (!parameter.getParameterSegmentationInputFile2().getMask().equals("")) {
referenceClusterSet = MainTools.readTheSecondClusterSet(parameter);
}
ClusterSet uemClusterSet = null;
if (!parameter.getParameterSegmentationInputFile3().getMask().equals("")) {
referenceClusterSet = MainTools.readThe3rdClusterSet(parameter);
}
ParameterBNDiarization parameterDiarization = parameter.getParameterDiarization();
// ** mask for the output of the segmentation file
ClusterSet clusterSet = initialize(parameter);
// ** load the features, sphinx format (13 MFCC with C0) or compute it form a wave file
AudioFeatureSet featureSet =
loadFeature(
parameter,
clusterSet,
parameter.getParameterInputFeature().getFeaturesDescriptorAsString());
featureSet.setCurrentShow(parameter.show);
int nbFeatures = featureSet.getNumberOfFeatures();
if (parameter.getParameterDiarization().isLoadInputSegmentation() == false) {
clusterSet.getFirstCluster().firstSegment().setLength(nbFeatures);
}
// clusterSet.debug(3);
ClusterSet clustersSegInit = sanityCheck(clusterSet, featureSet, parameter);
ClusterSet clustersSeg = segmentation("GLR", "FULL", clustersSegInit, featureSet, parameter);
// Seg IRIT
// ClusterSet clustersSegInit = clusterSet;
// ClusterSet clustersSeg = clusterSet;
// Seg IRIT
ClusterSet clustersLClust =
clusteringLinear(
parameterDiarization.getThreshold("l"), clustersSeg, featureSet, parameter);
ClusterSet clustersHClust =
clustering(parameterDiarization.getThreshold("h"), clustersLClust, featureSet, parameter);
// MainTools.writeClusterSet(parameter, clustersHClust, false);
ClusterSet clustersDClust =
decode(8, parameterDiarization.getThreshold("d"), clustersHClust, featureSet, parameter);
ClusterSet clustersSplitClust =
speech("10,10,50", clusterSet, clustersSegInit, clustersDClust, featureSet, parameter);
ClusterSet clustersGender = gender(clusterSet, clustersSplitClust, featureSet, parameter);
if (parameter.getParameterDiarization().isCEClustering()) {
ClusterSet clustersCLR =
speakerClustering(
parameterDiarization.getThreshold("c"),
"ce",
clustersSegInit,
clustersGender,
featureSet,
parameter);
MainTools.writeClusterSet(parameter, clustersCLR, false);
if (referenceClusterSet != null) {
DiarizationError computeError = new DiarizationError(referenceClusterSet, uemClusterSet);
computeError.scoreOfMatchedSpeakers(clustersCLR);
}
} else {
MainTools.writeClusterSet(parameter, clustersGender, false);
}
}