Created
April 21, 2019 18:18
-
-
Save anisayari/b64c172de1142a0e991068810deb2d3f to your computer and use it in GitHub Desktop.
extract_features_from_songs.py
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import librosa | |
import numpy as np | |
import pandas as pd | |
from tqdm import tqdm | |
def extract_features_from_a_song(x,sr): | |
dict_features = { | |
'zcr': np.mean(librosa.feature.zero_crossing_rate(x)), | |
'chroma_stft': np.mean(librosa.feature.chroma_stft(x, sr=sr)), | |
'mfcc': np.mean(librosa.feature.mfcc(x, sr=sr)), | |
'spec_cent': np.mean(librosa.feature.spectral_centroid(x, sr=sr)), | |
'spec_bw': np.mean(librosa.feature.spectral_bandwidth(x, sr=sr)), | |
'spec_rolloff': np.mean(librosa.feature.spectral_centroid(x, sr=sr)) | |
} | |
return dict_features | |
def get_data_from_sound(row): | |
dict_features = {} | |
dict_features['chroma_stft'] = 0 | |
dict_features['mfcc'] = 0 | |
dict_features['spec_cent'] = 0 | |
dict_features['spec_rolloff'] = 0 | |
dict_features['spec_bw'] = 0 | |
dict_features['zcr'] = 0 | |
audio_path = 'data/music/{}/{}.mp3'.format(row['style'],str(row['uuid'])) | |
exists = os.path.isfile(audio_path) | |
if exists : | |
x, sr = librosa.load(audio_path) | |
uuid = row['uuid'] | |
try: | |
pass | |
#create_waveform_image(x, sr, uuid) | |
except OverflowError: | |
print('Cannot save waveform file for',uuid) | |
try: | |
pass | |
#create_spectrogram_image(x,sr,uuid) | |
except OverflowError: | |
print('Cannot save spectogram file for',uuid) | |
try: | |
dict_features = extract_features_from_a_song(x,sr) | |
except OverflowError: | |
print('Cannot run features extract for',uuid) | |
print('{} DONE'.format(uuid)) | |
row['spec_cent'], row['spec_rolloff'], \ | |
row['spec_bw'], row['mfcc'], row['zcr'], row['chroma_stft'] = dict_features['spec_cent'], \ | |
dict_features['spec_rolloff'], \ | |
dict_features['spec_bw'], \ | |
dict_features['mfcc'], \ | |
dict_features['zcr'], \ | |
dict_features['chroma_stft'] | |
return row | |
tqdm.pandas() | |
df = pd.read_csv(output_file,sep=";", header=0) | |
df = df.progress_apply(get_data_from_sound , axis=1) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment