Datasets:
Tasks:
Audio Classification
Modalities:
Image
Formats:
imagefolder
Languages:
English
Size:
< 1K
Tags:
SER
Speech Emotion Recognition
Speech Emotion Classification
Audio Classification
Audio
Emotion
License:
File size: 1,029 Bytes
519bb97 |
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 |
# REQUIREMENTS
# $ pip install scipy
# $ pip install "polars[all]"
# You may need to install snappy in order to run this script:
# $ sudo pacman -S snappy
# $ pip install python-snappy
import polars as pl
import numpy as np
from scipy.io.wavfile import write
import os
n_cores = str(os.cpu_count())
os.environ['OMP_NUM_THREADS'] = n_cores
os.environ['MKL_NUM_THREADS'] = n_cores
if not os.path.isdir('wavs'):
os.makedirs('wavs')
columns = ['ytid', 'ytid_seg', 'start', 'end', 'sentiment', 'happiness', 'sadness', 'anger', 'fear', 'disgust', 'surprise']
df = pl.read_parquet('sqe_messai.parquet', columns = columns)
df.write_csv('sqe_messai_nowav.csv')
print(df)
columns2 = ['ytid_seg', 'wav2numpy']
df2 = pl.read_parquet('sqe_messai.parquet', use_pyarrow=False, columns = columns2)
def numpy2wav(row):
segment = os.path.splitext(os.path.basename(os.path.normpath(row[0])))[0]
print('PROCESSED:', segment)
write('wavs/' + segment + '.wav', 16000, np.array(row[1]))
return segment
df2.apply(lambda x: numpy2wav(x))
|