-
Notifications
You must be signed in to change notification settings - Fork 8
/
doPhonemeClassification.py
229 lines (176 loc) · 7.83 KB
/
doPhonemeClassification.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
#!/usr/bin/env python
# -*- coding: utf-8 -*-
'''
* Copyright (C) 2017 Music Technology Group - Universitat Pompeu Fabra
*
* This file is part of EUSIPCO2017 phoneme classification
*
* pypYIN is free software: you can redistribute it and/or modify it under
* the terms of the GNU Affero General Public License as published by the Free
* Software Foundation (FSF), either version 3 of the License, or (at your
* option) any later version.
*
* This program is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
* FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
* details.
*
* You should have received a copy of the Affero GNU General Public License
* version 3 along with this program. If not, see http://www.gnu.org/licenses/
*
* If you have any problem about this python version code, please contact: Rong Gong
*
*
* If you want to refer this code, please use this article:
*
'''
import os
import numpy as np
import pickle
import essentia.standard as ess
from sklearn import preprocessing
from parameters import *
from phonemeMap import *
from textgridParser import syllableTextgridExtraction
from trainTestSeparation import getRecordingNames
from phonemeSampleCollection import getFeature,getMFCCBands1D,getMFCCBands2D,featureReshape
from phonemeClassification import PhonemeClassification
from sklearn.metrics import confusion_matrix, accuracy_score
import matplotlib.pyplot as plt
def doClassification():
"""
1. collect features from test set
2. predict by GMM or DNN models
3. save the prediction
:return: prediction of GMM and DNN model
"""
phone_class = PhonemeClassification()
phone_class.create_gmm(gmmModel_path)
mfcc_all = np.array([])
mfccBands1D_all = np.array([])
mfccBands2D_all = np.array([])
y_true = []
for recording in getRecordingNames('TEST', dataset):
nestedPhonemeLists, numSyllables, numPhonemes \
= syllableTextgridExtraction(textgrid_path,recording,syllableTierName,phonemeTierName)
wav_full_filename = os.path.join(wav_path,recording+'.wav')
audio = ess.MonoLoader(downmix = 'left', filename = wav_full_filename, sampleRate = fs)()
# plotAudio(audio,15,16)
print 'calculating mfcc and mfcc bands ... ', recording
mfcc = getFeature(audio, d=True, nbf=False)
mfccBands1D = getMFCCBands1D(audio, nbf=True)
mfccBands2D = getMFCCBands2D(audio, nbf=True)
mfccBands2D = np.log(10000*mfccBands2D+1)
# scale mfccBands1D for dnn acoustic models
mfccBands1D_std = preprocessing.StandardScaler().fit_transform(mfccBands1D)
# scale mfccBands2D for cnn acoustic models
scaler = pickle.load(open(scaler_path, 'rb'))
mfccBands2D_std = scaler.transform(mfccBands2D)
for ii,pho in enumerate(nestedPhonemeLists):
print 'calculating ', recording, ' and phoneme ', str(ii), ' of ', str(len(nestedPhonemeLists))
# MFCC feature
sf = round(pho[0][0]*fs/hopsize)
ef = round(pho[0][1]*fs/hopsize)
# mfcc syllable
mfcc_s = mfcc[sf:ef,:]
mfccBands_s = mfccBands2D[sf:ef,:]
mfccBands1D_s_std = mfccBands1D_std[sf:ef,:]
mfccBands2D_s_std = mfccBands2D_std[sf:ef,:]
if len(mfcc_all):
mfcc_all = np.vstack((mfcc_all,mfcc_s))
mfccBands1D_all = np.vstack((mfccBands1D_all,mfccBands1D_s_std))
mfccBands2D_all = np.vstack((mfccBands2D_all,mfccBands2D_s_std))
else:
mfcc_all = mfcc_s
mfccBands1D_all = mfccBands1D_s_std
mfccBands2D_all = mfccBands2D_s_std
# print mfcc_all.shape, mfccBands2D_all.shape
##-- parsing y_true
y_true_s = []
for ii_p, p in enumerate(pho[1]):
# map from annotated xsampa to readable notation
key = dic_pho_map[p[2]]
index_key = dic_pho_label[key]
y_true_s += [index_key]*int(round((p[1]-p[0])/hopsize_t))
print len(y_true_s), mfcc_s.shape[0]
if len(y_true_s) > mfcc_s.shape[0]:
y_true_s = y_true_s[:mfcc_s.shape[0]]
elif len(y_true_s) < mfcc_s.shape[0]:
y_true_s += [y_true_s[-1]]*(mfcc_s.shape[0]-len(y_true_s))
y_true += y_true_s
phone_class.mapb_gmm(mfcc_all)
obs_gmm = phone_class.mapb_gmm_getter()
y_pred_gmm = phone_class.prediction(obs_gmm)
mfccBands2D_all = featureReshape(mfccBands2D_all)
phone_class.mapb_keras(mfccBands2D_all, kerasModels_jordi_path, jordi=True)
obs_cnn_jordi = phone_class.mapb_keras_getter()
y_pred_jordi = phone_class.prediction(obs_cnn_jordi)
phone_class.mapb_keras(mfccBands2D_all, kerasModels_choi_path)
obs_cnn_choi = phone_class.mapb_keras_getter()
y_pred_choi = phone_class.prediction(obs_cnn_choi)
phone_class.mapb_keras(mfccBands1D_all, kerasModels_dnn_path)
obs_dnn = phone_class.mapb_keras_getter()
y_pred_dnn = phone_class.prediction(obs_dnn)
np.save('./trainingData/y_pred_gmm.npy',y_pred_gmm)
np.save('./trainingData/y_pred_jordi.npy',y_pred_jordi)
np.save('./trainingData/y_pred_choi.npy',y_pred_choi)
np.save('./trainingData/y_pred_dnn.npy',y_pred_dnn)
np.save('./trainingData/y_true.npy',y_true)
def plot_confusion_matrix(cm, classes,
normalize=False,
title='Confusion matrix',
cmap=plt.cm.Blues):
"""
This function prints and plots the confusion matrix.
Normalization can be applied by setting `normalize=True`.
"""
if normalize:
cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]
print("Normalized confusion matrix")
else:
print('Confusion matrix, without normalization')
plt.imshow(cm, interpolation='nearest', cmap=cmap)
plt.title(title)
plt.colorbar()
tick_marks = np.arange(len(classes))
plt.xticks(tick_marks, classes, rotation=45)
plt.yticks(tick_marks, classes)
# print(cm)
plt.tight_layout()
plt.ylabel('True label')
plt.xlabel('Predicted label')
def draw_confusion_matrix(pred_path, true_path, title):
y_pred = np.load(pred_path)
y_true = np.load(true_path)
label = []
for ii in xrange(len(dic_pho_label_inv)):
label.append(dic_pho_label_inv[ii])
cm = confusion_matrix(y_true,y_pred)
plt.figure()
plot_confusion_matrix(cm, label,
normalize=True,
title= title +' model confusion matrix. accuracy: '
+ str(accuracy_score(y_true, y_pred)))
plt.show()
if __name__ == '__main__':
####---- predict the phoneme classes on test set
pred_gmm_path = './trainingData/y_pred_gmm.npy'
pred_jordi_path = './trainingData/y_pred_jordi.npy'
pred_choi_path = './trainingData/y_pred_choi.npy'
pred_dnn_path = './trainingData/y_pred_dnn.npy'
true_path = './trainingData/y_true.npy'
doClassification()
####---- draw the confusion matrix
draw_confusion_matrix(pred_path=pred_gmm_path,
true_path=true_path,
title='GMM')
draw_confusion_matrix(pred_path=pred_jordi_path,
true_path=true_path,
title='Proposed CNN')
draw_confusion_matrix(pred_path=pred_choi_path,
true_path=true_path,
title='Choi CNN')
draw_confusion_matrix(pred_path=pred_dnn_path,
true_path=true_path,
title='DNN')