def main_gen(params, filenum):
'''Calls gen_audio() to generate the audio signals, verifies that they meet
the requirements, and writes the files to storage'''
print("Generating file #" + str(filenum))
clean_clipped_files = []
clean_low_activity_files = []
noise_clipped_files = []
noise_low_activity_files = []
while True:
# generate clean speech
clean, clean_source_files, clean_cf, clean_laf = \
gen_audio(True, params, filenum)
# generate noise
noise, noise_source_files, noise_cf, noise_laf = \
gen_audio(False, params, filenum, len(clean))
clean_clipped_files += clean_cf
clean_low_activity_files += clean_laf
noise_clipped_files += noise_cf
noise_low_activity_files += noise_laf
# mix clean speech and noise
# if specified, use specified SNR value
if not params['randomize_snr']:
snr = params['snr']
# use a randomly sampled SNR value between the specified bounds
else:
snr = np.random.randint(params['snr_lower'], params['snr_upper'])
clean_snr, noise_snr, noisy_snr, target_level = snr_mixer(params=params,
clean=clean,
noise=noise,
snr=snr)
# Uncomment the below lines if you need segmental SNR and comment the above lines using snr_mixer
#clean_snr, noise_snr, noisy_snr, target_level = segmental_snr_mixer(params=params,
# clean=clean,
# noise=noise,
# snr=snr)
# unexpected clipping
if is_clipped(clean_snr) or is_clipped(noise_snr) or is_clipped(noisy_snr):
continue
else:
break
# write resultant audio streams to files
hyphen = '-'
clean_source_filenamesonly = [i[:-4].split(os.path.sep)[-1] for i in clean_source_files]
clean_files_joined = hyphen.join(clean_source_filenamesonly)[:MAXFILELEN]
noise_source_filenamesonly = [i[:-4].split(os.path.sep)[-1] for i in noise_source_files]
noise_files_joined = hyphen.join(noise_source_filenamesonly)[:MAXFILELEN]
noisyfilename = clean_files_joined + '_' + noise_files_joined + '_snr' + \
str(snr) + '_fileid_' + str(filenum) + '.wav'
cleanfilename = 'clean_fileid_'+str(filenum)+'.wav'
noisefilename = 'noise_fileid_'+str(filenum)+'.wav'
noisypath = os.path.join(params['noisyspeech_dir'], noisyfilename)
cleanpath = os.path.join(params['clean_proc_dir'], cleanfilename)
noisepath = os.path.join(params['noise_proc_dir'], noisefilename)
audio_signals = [noisy_snr, clean_snr, noise_snr]
file_paths = [noisypath, cleanpath, noisepath]
for i in range(len(audio_signals)):
try:
audiowrite(file_paths[i], audio_signals[i], params['fs'])
except Exception as e:
print(str(e))
pass
return clean_source_files, clean_clipped_files, clean_low_activity_files, \
noise_source_files, noise_clipped_files, noise_low_activity_files
def main(cfg):
snr_lower = int(cfg["snr_lower"])
snr_upper = int(cfg["snr_upper"])
total_snrlevels = int(cfg["total_snrlevels"])
clean_dir = os.path.join(os.path.dirname(__file__), 'clean_train')
if cfg["speech_dir"] != 'None':
clean_dir = cfg["speech_dir"]
if not os.path.exists(clean_dir):
assert False, ("Clean speech data is required")
try:
noise_dir = os.path.join(os.path.dirname(__file__), 'noise_train_new')
if cfg["noise_dir"] != 'None':
noise_dir = cfg["noise_dir"]
if not os.path.exists(noise_dir):
raise Exception("path does not exist")
if not len(os.listdir(noise_dir)):
raise Exception("folder is empty")
except:
res = input(
"Can't find noise data. Do you want to run the audio slicer?\nEnter [Y]es [N]o\n"
)
if res.lower() in ["y", "yes", ""]:
slicer()
print("Sliced successfully.\n\n")
else:
input("Noise data is required.\nPress any key to exit.\n")
sys.exit()
fs = float(cfg["sampling_rate"])
audioformat = cfg["audioformat"]
total_hours = float(cfg["total_hours"])
audio_length = float(cfg["audio_length"])
silence_length = float(cfg["silence_length"])
noisyspeech_dir = os.path.join(os.path.dirname(__file__), 'mix')
if not os.path.exists(noisyspeech_dir):
os.makedirs(noisyspeech_dir)
clean_proc_dir = os.path.join(os.path.dirname(__file__), 's1')
if not os.path.exists(clean_proc_dir):
os.makedirs(clean_proc_dir)
noise_proc_dir = os.path.join(os.path.dirname(__file__), 's2')
if not os.path.exists(noise_proc_dir):
os.makedirs(noise_proc_dir)
total_secs = total_hours * 60 * 60
total_samples = int(total_secs * fs)
audio_length = int(audio_length * fs)
SNR = np.linspace(snr_lower, snr_upper, total_snrlevels)
cleanfilenames = glob.glob(os.path.join(clean_dir, audioformat))
if cfg["noise_types_excluded"] == 'None':
noisefilenames = glob.glob(os.path.join(noise_dir, audioformat))
else:
filestoexclude = cfg["noise_types_excluded"].split(',')
noisefilenames = glob.glob(os.path.join(noise_dir, audioformat))
for i in range(len(filestoexclude)):
noisefilenames = [
fn for fn in noisefilenames
if not os.path.basename(fn).startswith(filestoexclude[i])
]
filecounter = 0
num_samples = 0
while num_samples < total_samples:
idx_s = np.random.randint(0, np.size(cleanfilenames))
clean, fs = audioread(cleanfilenames[idx_s])
if len(clean) == audio_length:
clean = clean
else:
while len(clean) <= audio_length:
idx_s = idx_s + 1
if idx_s >= np.size(cleanfilenames) - 1:
idx_s = np.random.randint(0, np.size(cleanfilenames))
newclean, fs = audioread(cleanfilenames[idx_s])
cleanconcat = np.append(clean,
np.zeros(int(fs * silence_length)))
clean = np.append(cleanconcat, newclean)
clean = clean[0:audio_length]
idx_n = np.random.randint(0, np.size(noisefilenames))
noise, fs = audioread(noisefilenames[idx_n])
if len(noise) >= len(clean):
noise = noise[0:len(clean)]
else:
while len(noise) <= len(clean):
idx_n = idx_n + 1
if idx_n >= np.size(noisefilenames) - 1:
idx_n = np.random.randint(0, np.size(noisefilenames))
newnoise, fs = audioread(noisefilenames[idx_n])
noiseconcat = np.append(noise,
np.zeros(int(fs * silence_length)))
noise = np.append(noiseconcat, newnoise)
noise = noise[0:len(clean)]
filecounter = filecounter + 1
for i in range(np.size(SNR)):
clean_snr, noise_snr, noisy_snr = snr_mixer(clean=clean,
noise=noise,
snr=SNR[i])
noisyfilename = 'noisy' + str(filecounter) + '_SNRdb_' + str(
SNR[i]) + noisefilenames[idx_n].split("\\")[-1].split(
".")[0] + '.wav'
cleanfilename = 'clnsp' + str(filecounter) + '_.wav'
noisefilename = 'noise' + str(filecounter) + '_SNRdb_' + str(
SNR[i]) + noisefilenames[idx_n].split("\\")[-1].split(
".")[0] + '.wav'
noisypath = os.path.join(noisyspeech_dir, noisyfilename)
cleanpath = os.path.join(clean_proc_dir, cleanfilename)
noisepath = os.path.join(noise_proc_dir, noisefilename)
audiowrite(noisy_snr, fs, noisypath, norm=False)
audiowrite(clean_snr, fs, cleanpath, norm=False)
audiowrite(noise_snr, fs, noisepath, norm=False)
num_samples = num_samples + len(noisy_snr)
def main(cfg):
snr_lower = float(cfg["snr_lower"])
snr_upper = float(cfg["snr_upper"])
total_snrlevels = float(cfg["total_snrlevels"])
clean_dir = os.path.join(os.path.dirname(__file__), 'clean_train')
if cfg["speech_dir"] != 'None':
clean_dir = cfg["speech_dir"]
if not os.path.exists(clean_dir):
assert False, ("Clean speech data is required")
noise_dir = os.path.join(os.path.dirname(__file__), 'noise_train')
if cfg["noise_dir"] != 'None':
noise_dir = cfg["noise_dir"]
if not os.path.exists(noise_dir):
assert False, ("Noise data is required")
fs = float(cfg["sampling_rate"])
audioformat = cfg["audioformat"]
total_hours = float(cfg["total_hours"])
audio_length = float(cfg["audio_length"])
silence_length = float(cfg["silence_length"])
noisyspeech_dir = os.path.join(os.path.dirname(__file__),
'NoisySpeech_training')
if not os.path.exists(noisyspeech_dir):
os.makedirs(noisyspeech_dir)
clean_proc_dir = os.path.join(os.path.dirname(__file__),
'CleanSpeech_training')
if not os.path.exists(clean_proc_dir):
os.makedirs(clean_proc_dir)
noise_proc_dir = os.path.join(os.path.dirname(__file__), 'Noise_training')
if not os.path.exists(noise_proc_dir):
os.makedirs(noise_proc_dir)
total_secs = total_hours * 60 * 60
total_samples = int(total_secs * fs)
audio_length = int(audio_length * fs)
SNR = np.linspace(snr_lower, snr_upper, total_snrlevels)
cleanfilenames = glob.glob(os.path.join(clean_dir, audioformat))
if cfg["noise_types_excluded"] == 'None':
noisefilenames = glob.glob(os.path.join(noise_dir, audioformat))
else:
filestoexclude = cfg["noise_types_excluded"].split(',')
noisefilenames = glob.glob(os.path.join(noise_dir, audioformat))
for i in range(len(filestoexclude)):
noisefilenames = [
fn for fn in noisefilenames
if not os.path.basename(fn).startswith(filestoexclude[i])
]
filecounter = 0
num_samples = 0
while num_samples < total_samples:
idx_s = np.random.randint(0, np.size(cleanfilenames))
clean, fs = audioread(cleanfilenames[idx_s])
if len(clean) > audio_length:
clean = clean
else:
while len(clean) <= audio_length:
idx_s = idx_s + 1
if idx_s >= np.size(cleanfilenames) - 1:
idx_s = np.random.randint(0, np.size(cleanfilenames))
newclean, fs = audioread(cleanfilenames[idx_s])
cleanconcat = np.append(clean,
np.zeros(int(fs * silence_length)))
clean = np.append(cleanconcat, newclean)
idx_n = np.random.randint(0, np.size(noisefilenames))
noise, fs = audioread(noisefilenames[idx_n])
if len(noise) >= len(clean):
noise = noise[0:len(clean)]
else:
while len(noise) <= len(clean):
idx_n = idx_n + 1
if idx_n >= np.size(noisefilenames) - 1:
idx_n = np.random.randint(0, np.size(noisefilenames))
newnoise, fs = audioread(noisefilenames[idx_n])
noiseconcat = np.append(noise,
np.zeros(int(fs * silence_length)))
noise = np.append(noiseconcat, newnoise)
noise = noise[0:len(clean)]
filecounter = filecounter + 1
for i in range(np.size(SNR)):
clean_snr, noise_snr, noisy_snr = snr_mixer(clean=clean,
noise=noise,
snr=SNR[i])
noisyfilename = 'noisy' + str(filecounter) + '_SNRdb_' + str(
SNR[i]) + '_clnsp' + str(filecounter) + '.wav'
cleanfilename = 'clnsp' + str(filecounter) + '.wav'
noisefilename = 'noisy' + str(filecounter) + '_SNRdb_' + str(
SNR[i]) + '.wav'
noisypath = os.path.join(noisyspeech_dir, noisyfilename)
cleanpath = os.path.join(clean_proc_dir, cleanfilename)
noisepath = os.path.join(noise_proc_dir, noisefilename)
audiowrite(noisy_snr, fs, noisypath, norm=False)
audiowrite(clean_snr, fs, cleanpath, norm=False)
audiowrite(noise_snr, fs, noisepath, norm=False)
num_samples = num_samples + len(noisy_snr)
from audiolib import audioread, audiowrite, snr_mixer
from os.path import isfile, join, basename
import os
import numpy as np
"""
Given a source folder, add white-noise in a range of different SNR levels to all files in the source folder.
"""
if __name__=="__main__":
source_folder = "clips"
source_files = [join(source_folder, f) for f in os.listdir(source_folder) if isfile(join(source_folder, f))]
output_folder = source_folder+"_snr"
snr_min = 30
snr_max = 50
for f in source_files:
clean, fs = audioread(f)
# white-noise
noise = np.random.normal(0, 1, len(clean))
for i in range (snr_min,snr_max):
clean_snr, noise_snr, noisy_snr = snr_mixer(clean=clean, noise=noise, snr=i)
output_filename = join(output_folder,
f'{i}S_{os.path.splitext(basename(f))[0]}.wav')
audiowrite(noisy_snr, fs,output_filename, norm=False)