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

seprated spectrogram generation code

master
Malar Kannan 2017-10-17 19:11:04 +05:30
parent 51a6d6e804
commit 88edcdd239
1 changed files with 10 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
speech_data.py
View File

@ -56,20 +56,23 @@ def sunflower_pairs_data():
te_pairs = np.array([x_pos_test,x_neg_test]).reshape(x_pos_test.shape[0],2,max_samples,sample_size) te_pairs = np.array([x_pos_test,x_neg_test]).reshape(x_pos_test.shape[0],2,max_samples,sample_size)
return tr_pairs,te_pairs,tr_y,te_y return tr_pairs,te_pairs,tr_y,te_y
def create_spectrogram_data():
audio_samples = pd.read_csv('./outputs/'+audio_group+'.csv',names=['word','voice','rate','variant','file'])
audio_samples.loc[:,'spectrogram'] = audio_samples.loc[:,'file'].apply(lambda x:'outputs/'+audio_group+'/'+x).apply(generate_aiff_spectrogram)
audio_samples.to_pickle('spectrogram.pkl')
def speech_pairs_data(audio_group): def speech_pairs_data(audio_group):
audio_samples = pd.read_csv('./outputs/'+audio_group+'.csv',names=['word','voice','rate','variant','file']) audio_samples = pd.read_pickle('spectrogram.pkl')
audio_samples.loc[:,'file'] = audio_samples.loc[:,'file'].apply(lambda x:'outputs/'+audio_group+'/'+x).apply(generate_aiff_spectrogram)
y_data = audio_samples['variant'].apply(lambda x:x=='normal').values y_data = audio_samples['variant'].apply(lambda x:x=='normal').values
max_samples = audio_samples['file'].apply(lambda x:x.shape[0]).max() max_samples = audio_samples['spectrogram'].apply(lambda x:x.shape[0]).max()
sample_size = audio_samples['file'][0].shape[1] sample_size = audio_samples['spectrogram'][0].shape[1]
audio_samples_pos = audio_samples[audio_samples['variant'] == 'normal'].reset_index(drop=True) audio_samples_pos = audio_samples[audio_samples['variant'] == 'normal'].reset_index(drop=True)
audio_samples_neg = audio_samples[audio_samples['variant'] == 'phoneme'].reset_index(drop=True) audio_samples_neg = audio_samples[audio_samples['variant'] == 'phoneme'].reset_index(drop=True)
def append_zeros(spgr): def append_zeros(spgr):
return np.lib.pad(spgr,[(0, max_samples-spgr.shape[0]), (0,0)],'median') return np.lib.pad(spgr,[(0, max_samples-spgr.shape[0]), (0,0)],'median')
def create_data(sf): def create_data(sf):
sample_count = sf['file'].shape[0] sample_count = sf['spectrogram'].shape[0]
pad_sun = sf['file'].apply(append_zeros).values pad_sun = sf['spectrogram'].apply(append_zeros).values
x_data = np.vstack(pad_sun).reshape((sample_count,max_samples,sample_size)) x_data = np.vstack(pad_sun).reshape((sample_count,max_samples,sample_size))
return x_data return x_data
x_data_pos = create_data(audio_samples_pos) x_data_pos = create_data(audio_samples_pos)
@ -82,4 +85,5 @@ def speech_pairs_data(audio_group):
return tr_pairs,te_pairs,tr_y,te_y return tr_pairs,te_pairs,tr_y,te_y
if __name__ == '__main__': if __name__ == '__main__':
create_spectrogram_data()
print(speech_pairs_data()) print(speech_pairs_data())