Spaces:
Sleeping
Sleeping
Upload trainset_preprocess_pipeline_print.py
Browse files
trainset_preprocess_pipeline_print.py
ADDED
|
@@ -0,0 +1,135 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import sys, os, multiprocessing
|
| 2 |
+
from scipy import signal
|
| 3 |
+
|
| 4 |
+
now_dir = os.getcwd()
|
| 5 |
+
sys.path.append(now_dir)
|
| 6 |
+
|
| 7 |
+
inp_root = sys.argv[1]
|
| 8 |
+
sr = int(sys.argv[2])
|
| 9 |
+
n_p = int(sys.argv[3])
|
| 10 |
+
exp_dir = sys.argv[4]
|
| 11 |
+
noparallel = sys.argv[5] == "True"
|
| 12 |
+
import numpy as np, os, traceback
|
| 13 |
+
from slicer2 import Slicer
|
| 14 |
+
import librosa, traceback
|
| 15 |
+
from scipy.io import wavfile
|
| 16 |
+
import multiprocessing
|
| 17 |
+
from my_utils import load_audio
|
| 18 |
+
|
| 19 |
+
mutex = multiprocessing.Lock()
|
| 20 |
+
f = open("%s/preprocess.log" % exp_dir, "a+")
|
| 21 |
+
|
| 22 |
+
|
| 23 |
+
def println(strr):
|
| 24 |
+
mutex.acquire()
|
| 25 |
+
print(strr)
|
| 26 |
+
f.write("%s\n" % strr)
|
| 27 |
+
f.flush()
|
| 28 |
+
mutex.release()
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
class PreProcess:
|
| 32 |
+
def __init__(self, sr, exp_dir):
|
| 33 |
+
self.slicer = Slicer(
|
| 34 |
+
sr=sr,
|
| 35 |
+
threshold=-40,
|
| 36 |
+
min_length=800,
|
| 37 |
+
min_interval=400,
|
| 38 |
+
hop_size=15,
|
| 39 |
+
max_sil_kept=150,
|
| 40 |
+
)
|
| 41 |
+
self.sr = sr
|
| 42 |
+
self.bh, self.ah = signal.butter(N=5, Wn=48, btype="high", fs=self.sr)
|
| 43 |
+
self.per = 3.7
|
| 44 |
+
self.overlap = 0.3
|
| 45 |
+
self.tail = self.per + self.overlap
|
| 46 |
+
self.max = 0.95
|
| 47 |
+
self.alpha = 0.8
|
| 48 |
+
self.exp_dir = exp_dir
|
| 49 |
+
self.gt_wavs_dir = "%s/0_gt_wavs" % exp_dir
|
| 50 |
+
self.wavs16k_dir = "%s/1_16k_wavs" % exp_dir
|
| 51 |
+
os.makedirs(self.exp_dir, exist_ok=True)
|
| 52 |
+
os.makedirs(self.gt_wavs_dir, exist_ok=True)
|
| 53 |
+
os.makedirs(self.wavs16k_dir, exist_ok=True)
|
| 54 |
+
|
| 55 |
+
def norm_write(self, tmp_audio, idx0, idx1):
|
| 56 |
+
tmp_audio = (tmp_audio / np.abs(tmp_audio).max() * (self.max * self.alpha)) + (
|
| 57 |
+
1 - self.alpha
|
| 58 |
+
) * tmp_audio
|
| 59 |
+
wavfile.write(
|
| 60 |
+
"%s/%s_%s.wav" % (self.gt_wavs_dir, idx0, idx1),
|
| 61 |
+
self.sr,
|
| 62 |
+
tmp_audio.astype(np.float32),
|
| 63 |
+
)
|
| 64 |
+
tmp_audio = librosa.resample(
|
| 65 |
+
tmp_audio, orig_sr=self.sr, target_sr=16000
|
| 66 |
+
) # , res_type="soxr_vhq"
|
| 67 |
+
wavfile.write(
|
| 68 |
+
"%s/%s_%s.wav" % (self.wavs16k_dir, idx0, idx1),
|
| 69 |
+
16000,
|
| 70 |
+
tmp_audio.astype(np.float32),
|
| 71 |
+
)
|
| 72 |
+
|
| 73 |
+
def pipeline(self, path, idx0):
|
| 74 |
+
try:
|
| 75 |
+
audio = load_audio(path, self.sr)
|
| 76 |
+
# zero phased digital filter cause pre-ringing noise...
|
| 77 |
+
# audio = signal.filtfilt(self.bh, self.ah, audio)
|
| 78 |
+
audio = signal.lfilter(self.bh, self.ah, audio)
|
| 79 |
+
|
| 80 |
+
idx1 = 0
|
| 81 |
+
for audio in self.slicer.slice(audio):
|
| 82 |
+
i = 0
|
| 83 |
+
while 1:
|
| 84 |
+
start = int(self.sr * (self.per - self.overlap) * i)
|
| 85 |
+
i += 1
|
| 86 |
+
if len(audio[start:]) > self.tail * self.sr:
|
| 87 |
+
tmp_audio = audio[start : start + int(self.per * self.sr)]
|
| 88 |
+
self.norm_write(tmp_audio, idx0, idx1)
|
| 89 |
+
idx1 += 1
|
| 90 |
+
else:
|
| 91 |
+
tmp_audio = audio[start:]
|
| 92 |
+
idx1 += 1
|
| 93 |
+
break
|
| 94 |
+
self.norm_write(tmp_audio, idx0, idx1)
|
| 95 |
+
println("%s->Suc." % path)
|
| 96 |
+
except:
|
| 97 |
+
println("%s->%s" % (path, traceback.format_exc()))
|
| 98 |
+
|
| 99 |
+
def pipeline_mp(self, infos):
|
| 100 |
+
for path, idx0 in infos:
|
| 101 |
+
self.pipeline(path, idx0)
|
| 102 |
+
|
| 103 |
+
def pipeline_mp_inp_dir(self, inp_root, n_p):
|
| 104 |
+
try:
|
| 105 |
+
infos = [
|
| 106 |
+
("%s/%s" % (inp_root, name), idx)
|
| 107 |
+
for idx, name in enumerate(sorted(list(os.listdir(inp_root))))
|
| 108 |
+
]
|
| 109 |
+
if noparallel:
|
| 110 |
+
for i in range(n_p):
|
| 111 |
+
self.pipeline_mp(infos[i::n_p])
|
| 112 |
+
else:
|
| 113 |
+
ps = []
|
| 114 |
+
for i in range(n_p):
|
| 115 |
+
p = multiprocessing.Process(
|
| 116 |
+
target=self.pipeline_mp, args=(infos[i::n_p],)
|
| 117 |
+
)
|
| 118 |
+
p.start()
|
| 119 |
+
ps.append(p)
|
| 120 |
+
for p in ps:
|
| 121 |
+
p.join()
|
| 122 |
+
except:
|
| 123 |
+
println("Fail. %s" % traceback.format_exc())
|
| 124 |
+
|
| 125 |
+
|
| 126 |
+
def preprocess_trainset(inp_root, sr, n_p, exp_dir):
|
| 127 |
+
pp = PreProcess(sr, exp_dir)
|
| 128 |
+
println("start preprocess")
|
| 129 |
+
println(sys.argv)
|
| 130 |
+
pp.pipeline_mp_inp_dir(inp_root, n_p)
|
| 131 |
+
println("end preprocess")
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
if __name__ == "__main__":
|
| 135 |
+
preprocess_trainset(inp_root, sr, n_p, exp_dir)
|