malar
/
speech-scoring
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

1. fixed missing wrong pairs 

2.using different progress bakend
master
Malar Kannan 2017-11-07 17:27:01 +05:30
parent b8a9f87031
commit 7cbfebbf1a
2 changed files with 24 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
speech_data.py
View File

@ -11,11 +11,12 @@ import os
import random import random
import csv import csv
import gc import gc
import progressbar # import progressbar
from tqdm import tqdm
def prog_bar(title): # def prog_bar(title):
widgets = [title, progressbar.Counter(), ' [', progressbar.Bar(), '] - ', progressbar.ETA()] # widgets = [title, progressbar.Counter(), ' [', progressbar.Bar(), '] - ', progressbar.ETA()]
return progressbar.ProgressBar(widgets=widgets) # return progressbar.ProgressBar(widgets=widgets)
def siamese_pairs(rightGroup, wrongGroup): def siamese_pairs(rightGroup, wrongGroup):
group1 = [r for (i, r) in rightGroup.iterrows()] group1 = [r for (i, r) in rightGroup.iterrows()]
@ -26,7 +27,7 @@ def siamese_pairs(rightGroup, wrongGroup):
random.shuffle(rightRightPairs) random.shuffle(rightRightPairs)
# return (random.sample(same,10), random.sample(diff,10)) # return (random.sample(same,10), random.sample(diff,10))
# return rightRightPairs[:10],rightWrongPairs[:10] # return rightRightPairs[:10],rightWrongPairs[:10]
return rightRightPairs[:16],rightWrongPairs[:16] return rightRightPairs[:32],rightWrongPairs[:32]
# return rightRightPairs,rightWrongPairs # return rightRightPairs,rightWrongPairs
def create_spectrogram_tfrecords(audio_group='audio'): def create_spectrogram_tfrecords(audio_group='audio'):
@ -55,16 +56,21 @@ def create_spectrogram_tfrecords(audio_group='audio'):
return tf.train.Feature(bytes_list=tf.train.BytesList(value=value)) return tf.train.Feature(bytes_list=tf.train.BytesList(value=value))
writer = tf.python_io.TFRecordWriter('./outputs/' + audio_group + '.tfrecords') writer = tf.python_io.TFRecordWriter('./outputs/' + audio_group + '.tfrecords')
prog = prog_bar('Generating siamese pairs : ') prog = tqdm(audio_samples.groupby(audio_samples['word']),desc='Computing spectrogram')
for (w, word_group) in prog(audio_samples.groupby(audio_samples['word'])): for (w, word_group) in prog:
prog.set_postfix(word=w)
g = word_group.reset_index() g = word_group.reset_index()
g['spectrogram'] = apply_by_multiprocessing(g['file_path'],generate_aiff_spectrogram) g['spectrogram'] = apply_by_multiprocessing(g['file_path'],generate_aiff_spectrogram)
sample_right = g.loc[audio_samples['variant'] == 'low'] sample_right = g.loc[g['variant'] == 'low']
sample_wrong = g.loc[audio_samples['variant'] == 'medium'] sample_wrong = g.loc[g['variant'] == 'medium']
same, diff = siamese_pairs(sample_right, sample_wrong) same, diff = siamese_pairs(sample_right, sample_wrong)
groups = [([0,1],same),([1,0],diff)] groups = [([0,1],same),([1,0],diff)]
for (output,group) in groups: for (output,group) in groups:
for sample1,sample2 in group: group_prog = tqdm(group,desc='Writing Spectrogram')
for sample1,sample2 in group_prog:
group_prog.set_postfix(output=output
,var1=sample1['variant']
,var2=sample2['variant'])
spectro1,spectro2 = sample1['spectrogram'],sample2['spectrogram'] spectro1,spectro2 = sample1['spectrogram'],sample2['spectrogram']
spec_n1,spec_n2 = spectro1.shape[0],spectro2.shape[0] spec_n1,spec_n2 = spectro1.shape[0],spectro2.shape[0]
spec_w1,spec_w2 = spectro1.shape[1],spectro2.shape[1] spec_w1,spec_w2 = spectro1.shape[1],spectro2.shape[1]
@ -93,6 +99,7 @@ def create_spectrogram_tfrecords(audio_group='audio'):
} }
)) ))
writer.write(example.SerializeToString()) writer.write(example.SerializeToString())
prog.close()
writer.close() writer.close()
def padd_zeros(spgr, max_samples): def padd_zeros(spgr, max_samples):
@ -141,7 +148,7 @@ def read_siamese_tfrecords(audio_group='audio'):
# if len(input_pairs) > 50: # if len(input_pairs) > 50:
# break # break
input_data,output_data = np.asarray(input_pairs),np.asarray(output_class) input_data,output_data = np.asarray(input_pairs),np.asarray(output_class)
# import pdb; pdb.set_trace() import pdb; pdb.set_trace()
# tr_x1,te_x1,tr_x2,te_x2,tr_y,te_y = train_test_split(input1,input2,output_class) # tr_x1,te_x1,tr_x2,te_x2,tr_y,te_y = train_test_split(input1,input2,output_class)
tr_pairs,te_pairs,tr_y,te_y = train_test_split(input_data,output_data) tr_pairs,te_pairs,tr_y,te_y = train_test_split(input_data,output_data)
# return (tr_x1,te_x1,tr_x2,te_x2,tr_y,te_y) # return (tr_x1,te_x1,tr_x2,te_x2,tr_y,te_y)
@ -186,9 +193,9 @@ if __name__ == '__main__':
# sunflower_pairs_data() # sunflower_pairs_data()
# create_spectrogram_data() # create_spectrogram_data()
# create_spectrogram_data('story_words') # create_spectrogram_data('story_words')
# create_spectrogram_tfrecords('story_words') create_spectrogram_tfrecords('story_words')
# create_spectrogram_tfrecords('story_all') # create_spectrogram_tfrecords('story_words_test')
read_siamese_tfrecords('story_all') # read_siamese_tfrecords('story_all')
# create_padded_spectrogram() # create_padded_spectrogram()
# create_speech_pairs_data() # create_speech_pairs_data()
# print(speech_model_data()) # print(speech_model_data())

6
speech_siamese.py
View File

@ -114,11 +114,11 @@ def train_siamese():
rms = RMSprop(lr=0.001) rms = RMSprop(lr=0.001)
model.compile(loss=categorical_crossentropy, optimizer=rms, metrics=[accuracy]) model.compile(loss=categorical_crossentropy, optimizer=rms, metrics=[accuracy])
model.fit( model.fit(
[tr_x1, tr_x2], [tr_pairs[:, 0], tr_pairs[:, 1]],
tr_y, tr_y,
batch_size=128, batch_size=128,
epochs=50, epochs=100,
validation_data=([tr_pairs[:, 0], tr_pairs[:, 1]], te_y), validation_data=([te_pairs[:, 0], te_pairs[:, 1]], te_y),
callbacks=[tb_cb, cp_cb]) callbacks=[tb_cb, cp_cb])
model.save('./models/siamese_speech_model-final.h5') model.save('./models/siamese_speech_model-final.h5')