Skip to content
Snippets Groups Projects
Select Git revision
  • 5800a6f34225ecc0a4f98853fe4033e8d81337e6
  • master default protected
  • loss
  • producer
4 results

WordEmbeddings.hpp

Blame
    • get_spectrogram.py 2.75 KiB
    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 
    

    

    

    import os
import pandas as pd
import librosa
import numpy as np
import matplotlib.pyplot as plt
from tqdm import tqdm
import ipdb
import argparse
from p_tqdm import p_map
import warnings
warnings.filterwarnings('ignore')

parser = argparse.ArgumentParser(formatter_class=argparse.ArgumentDefaultsHelpFormatter, description='TODO')
parser.add_argument('-f','--file', type=str,required=True,help = 'Name of the file that contain the recording to print')
parser.add_argument('-p','--path_to_data', type=str, help = 'Path of the folder that contain the recordings',required=True)
parser.add_argument('-d','--direction', type=str, help = 'Directory to wich spectrogram will be stored',required=True)
parser.add_argument('-m','--mode', type=str,choices=['unique','multiple'], help = 'Direction of the saved spectrogram',required=True)
args = parser.parse_args()

path_to_data = args.path_to_data
direction = args.direction
folder = 'Spectrogram/'
DURATION = 5
OVERLAP = 2
if args.mode == 'multiple':
    NB_IMG_PER_REC = 10 
    
elif args.mode == 'unique': 
    NB_IMG_PER_REC = 1


df = pd.read_csv(str(args.file),low_memory=False)
df['Path'] = ''
#for w in range (len(df)):
#    df.loc[w,'Path']=str(args.path_to_data+ df.code_unique[w]+"_split_" + str(df.chunk_initial_time[w]) + "_" + str(df.chunk_final_time[w]))

for w in range (len(df)):
    df.loc[w,'Path']=str('/nfs/NAS7/Quebec2'+df.dir[w])

def process(x):
    _, (i) = x

    for count, j in enumerate(range (NB_IMG_PER_REC)): #30*8 secondes - 30*2 secondes (overlap) = 180 secondes affichées sur 30 images : n'affiche que les 3 premières minutes d'un enregistrement
    
        filename = str(i[0])
        offset = count * (DURATION - OVERLAP)
        try:
            y, sr = librosa.load(filename, offset = offset, duration = DURATION)
        except Exception:
            print(filename)
            continue
        window_size = 1024
        window = np.hanning(window_size)
        stft  = librosa.core.spectrum.stft(y, n_fft=window_size, hop_length=512, window=window)

        plt.close()
        plt.figure()

        vmin = np.flipud(np.log10(np.abs(stft))).mean()
        vmax = np.flipud(np.log10(np.abs(stft))).max()

        plt.imshow(np.flipud(np.log10(np.abs(stft))),aspect = "auto",interpolation = None,cmap = 'jet',vmin = vmin,vmax = vmax)

        plt.subplots_adjust(top=1, bottom=0, left=0, right=1)

        name = str(i[0].replace('/','_').split('.')[0]+'_'+str(count))
        try :
            plt.savefig(str(direction+folder+name+'.jpg'))

        except FileNotFoundError:

            os.mkdir(str(direction+folder))
            plt.savefig(str(direction+folder+name+'.jpg'))
        

p_map(process, enumerate(df.groupby('Path')), num_cpus=10) #la colonne dir correspond au path
print('saved to ',str(direction+folder))