Spaces:
Running
Running
Update vc_infer_pipeline.py
Browse files- vc_infer_pipeline.py +90 -32
vc_infer_pipeline.py
CHANGED
@@ -1,4 +1,4 @@
|
|
1 |
-
import numpy as np, parselmouth, torch, pdb
|
2 |
from time import time as ttime
|
3 |
import torch.nn.functional as F
|
4 |
import scipy.signal as signal
|
@@ -6,13 +6,17 @@ import pyworld, os, traceback, faiss, librosa, torchcrepe
|
|
6 |
from scipy import signal
|
7 |
from functools import lru_cache
|
8 |
|
|
|
|
|
|
|
9 |
bh, ah = signal.butter(N=5, Wn=48, btype="high", fs=16000)
|
10 |
|
11 |
-
input_audio_path2wav={}
|
|
|
12 |
|
13 |
@lru_cache
|
14 |
-
def cache_harvest_f0(input_audio_path,fs,f0max,f0min,frame_period):
|
15 |
-
audio=input_audio_path2wav[input_audio_path]
|
16 |
f0, t = pyworld.harvest(
|
17 |
audio,
|
18 |
fs=fs,
|
@@ -23,18 +27,29 @@ def cache_harvest_f0(input_audio_path,fs,f0max,f0min,frame_period):
|
|
23 |
f0 = pyworld.stonemask(audio, f0, t, fs)
|
24 |
return f0
|
25 |
|
26 |
-
|
|
|
27 |
# print(data1.max(),data2.max())
|
28 |
-
rms1 = librosa.feature.rms(
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
return data2
|
37 |
|
|
|
38 |
class VC(object):
|
39 |
def __init__(self, tgt_sr, config):
|
40 |
self.x_pad, self.x_query, self.x_center, self.x_max, self.is_half = (
|
@@ -54,7 +69,16 @@ class VC(object):
|
|
54 |
self.t_max = self.sr * self.x_max # 免查询时长阈值
|
55 |
self.device = config.device
|
56 |
|
57 |
-
def get_f0(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
58 |
global input_audio_path2wav
|
59 |
time_step = self.window / self.sr * 1000
|
60 |
f0_min = 50
|
@@ -78,9 +102,9 @@ class VC(object):
|
|
78 |
f0, [[pad_size, p_len - len(f0) - pad_size]], mode="constant"
|
79 |
)
|
80 |
elif f0_method == "harvest":
|
81 |
-
input_audio_path2wav[input_audio_path]=x.astype(np.double)
|
82 |
-
f0=cache_harvest_f0(input_audio_path,self.sr,f0_max,f0_min,10)
|
83 |
-
if
|
84 |
f0 = signal.medfilt(f0, 3)
|
85 |
elif f0_method == "crepe":
|
86 |
model = "full"
|
@@ -103,6 +127,15 @@ class VC(object):
|
|
103 |
f0 = torchcrepe.filter.mean(f0, 3)
|
104 |
f0[pd < 0.1] = 0
|
105 |
f0 = f0[0].cpu().numpy()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
106 |
f0 *= pow(2, f0_up_key / 12)
|
107 |
# with open("test.txt","w")as f:f.write("\n".join([str(i)for i in f0.tolist()]))
|
108 |
tf0 = self.sr // self.window # 每秒f0点数
|
@@ -125,7 +158,7 @@ class VC(object):
|
|
125 |
) + 1
|
126 |
f0_mel[f0_mel <= 1] = 1
|
127 |
f0_mel[f0_mel > 255] = 255
|
128 |
-
f0_coarse = np.rint(f0_mel).astype(int)
|
129 |
return f0_coarse, f0bak # 1-0
|
130 |
|
131 |
def vc(
|
@@ -141,6 +174,7 @@ class VC(object):
|
|
141 |
big_npy,
|
142 |
index_rate,
|
143 |
version,
|
|
|
144 |
): # ,file_index,file_big_npy
|
145 |
feats = torch.from_numpy(audio0)
|
146 |
if self.is_half:
|
@@ -161,8 +195,9 @@ class VC(object):
|
|
161 |
t0 = ttime()
|
162 |
with torch.no_grad():
|
163 |
logits = model.extract_features(**inputs)
|
164 |
-
feats = model.final_proj(logits[0])if version=="v1"else logits[0]
|
165 |
-
|
|
|
166 |
if (
|
167 |
isinstance(index, type(None)) == False
|
168 |
and isinstance(big_npy, type(None)) == False
|
@@ -188,6 +223,10 @@ class VC(object):
|
|
188 |
)
|
189 |
|
190 |
feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
|
|
|
|
|
|
|
|
|
191 |
t1 = ttime()
|
192 |
p_len = audio0.shape[0] // self.window
|
193 |
if feats.shape[1] < p_len:
|
@@ -195,6 +234,14 @@ class VC(object):
|
|
195 |
if pitch != None and pitchf != None:
|
196 |
pitch = pitch[:, :p_len]
|
197 |
pitchf = pitchf[:, :p_len]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
198 |
p_len = torch.tensor([p_len], device=self.device).long()
|
199 |
with torch.no_grad():
|
200 |
if pitch != None and pitchf != None:
|
@@ -206,10 +253,7 @@ class VC(object):
|
|
206 |
)
|
207 |
else:
|
208 |
audio1 = (
|
209 |
-
(net_g.infer(feats, p_len, sid)[0][0, 0])
|
210 |
-
.data.cpu()
|
211 |
-
.float()
|
212 |
-
.numpy()
|
213 |
)
|
214 |
del feats, p_len, padding_mask
|
215 |
if torch.cuda.is_available():
|
@@ -238,6 +282,7 @@ class VC(object):
|
|
238 |
resample_sr,
|
239 |
rms_mix_rate,
|
240 |
version,
|
|
|
241 |
f0_file=None,
|
242 |
):
|
243 |
if (
|
@@ -292,7 +337,15 @@ class VC(object):
|
|
292 |
sid = torch.tensor(sid, device=self.device).unsqueeze(0).long()
|
293 |
pitch, pitchf = None, None
|
294 |
if if_f0 == 1:
|
295 |
-
pitch, pitchf = self.get_f0(
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
296 |
pitch = pitch[:p_len]
|
297 |
pitchf = pitchf[:p_len]
|
298 |
if self.device == "mps":
|
@@ -317,6 +370,7 @@ class VC(object):
|
|
317 |
big_npy,
|
318 |
index_rate,
|
319 |
version,
|
|
|
320 |
)[self.t_pad_tgt : -self.t_pad_tgt]
|
321 |
)
|
322 |
else:
|
@@ -333,6 +387,7 @@ class VC(object):
|
|
333 |
big_npy,
|
334 |
index_rate,
|
335 |
version,
|
|
|
336 |
)[self.t_pad_tgt : -self.t_pad_tgt]
|
337 |
)
|
338 |
s = t
|
@@ -350,6 +405,7 @@ class VC(object):
|
|
350 |
big_npy,
|
351 |
index_rate,
|
352 |
version,
|
|
|
353 |
)[self.t_pad_tgt : -self.t_pad_tgt]
|
354 |
)
|
355 |
else:
|
@@ -366,19 +422,21 @@ class VC(object):
|
|
366 |
big_npy,
|
367 |
index_rate,
|
368 |
version,
|
|
|
369 |
)[self.t_pad_tgt : -self.t_pad_tgt]
|
370 |
)
|
371 |
audio_opt = np.concatenate(audio_opt)
|
372 |
-
if
|
373 |
-
audio_opt=change_rms(audio,16000,audio_opt,tgt_sr,rms_mix_rate)
|
374 |
-
if
|
375 |
audio_opt = librosa.resample(
|
376 |
audio_opt, orig_sr=tgt_sr, target_sr=resample_sr
|
377 |
)
|
378 |
-
audio_max=np.abs(audio_opt).max()/0.99
|
379 |
-
max_int16=32768
|
380 |
-
if
|
381 |
-
|
|
|
382 |
del pitch, pitchf, sid
|
383 |
if torch.cuda.is_available():
|
384 |
torch.cuda.empty_cache()
|
|
|
1 |
+
import numpy as np, parselmouth, torch, pdb, sys, os
|
2 |
from time import time as ttime
|
3 |
import torch.nn.functional as F
|
4 |
import scipy.signal as signal
|
|
|
6 |
from scipy import signal
|
7 |
from functools import lru_cache
|
8 |
|
9 |
+
now_dir = os.getcwd()
|
10 |
+
sys.path.append(now_dir)
|
11 |
+
|
12 |
bh, ah = signal.butter(N=5, Wn=48, btype="high", fs=16000)
|
13 |
|
14 |
+
input_audio_path2wav = {}
|
15 |
+
|
16 |
|
17 |
@lru_cache
|
18 |
+
def cache_harvest_f0(input_audio_path, fs, f0max, f0min, frame_period):
|
19 |
+
audio = input_audio_path2wav[input_audio_path]
|
20 |
f0, t = pyworld.harvest(
|
21 |
audio,
|
22 |
fs=fs,
|
|
|
27 |
f0 = pyworld.stonemask(audio, f0, t, fs)
|
28 |
return f0
|
29 |
|
30 |
+
|
31 |
+
def change_rms(data1, sr1, data2, sr2, rate): # 1是输入音频,2是输出音频,rate是2的占比
|
32 |
# print(data1.max(),data2.max())
|
33 |
+
rms1 = librosa.feature.rms(
|
34 |
+
y=data1, frame_length=sr1 // 2 * 2, hop_length=sr1 // 2
|
35 |
+
) # 每半秒一个点
|
36 |
+
rms2 = librosa.feature.rms(y=data2, frame_length=sr2 // 2 * 2, hop_length=sr2 // 2)
|
37 |
+
rms1 = torch.from_numpy(rms1)
|
38 |
+
rms1 = F.interpolate(
|
39 |
+
rms1.unsqueeze(0), size=data2.shape[0], mode="linear"
|
40 |
+
).squeeze()
|
41 |
+
rms2 = torch.from_numpy(rms2)
|
42 |
+
rms2 = F.interpolate(
|
43 |
+
rms2.unsqueeze(0), size=data2.shape[0], mode="linear"
|
44 |
+
).squeeze()
|
45 |
+
rms2 = torch.max(rms2, torch.zeros_like(rms2) + 1e-6)
|
46 |
+
data2 *= (
|
47 |
+
torch.pow(rms1, torch.tensor(1 - rate))
|
48 |
+
* torch.pow(rms2, torch.tensor(rate - 1))
|
49 |
+
).numpy()
|
50 |
return data2
|
51 |
|
52 |
+
|
53 |
class VC(object):
|
54 |
def __init__(self, tgt_sr, config):
|
55 |
self.x_pad, self.x_query, self.x_center, self.x_max, self.is_half = (
|
|
|
69 |
self.t_max = self.sr * self.x_max # 免查询时长阈值
|
70 |
self.device = config.device
|
71 |
|
72 |
+
def get_f0(
|
73 |
+
self,
|
74 |
+
input_audio_path,
|
75 |
+
x,
|
76 |
+
p_len,
|
77 |
+
f0_up_key,
|
78 |
+
f0_method,
|
79 |
+
filter_radius,
|
80 |
+
inp_f0=None,
|
81 |
+
):
|
82 |
global input_audio_path2wav
|
83 |
time_step = self.window / self.sr * 1000
|
84 |
f0_min = 50
|
|
|
102 |
f0, [[pad_size, p_len - len(f0) - pad_size]], mode="constant"
|
103 |
)
|
104 |
elif f0_method == "harvest":
|
105 |
+
input_audio_path2wav[input_audio_path] = x.astype(np.double)
|
106 |
+
f0 = cache_harvest_f0(input_audio_path, self.sr, f0_max, f0_min, 10)
|
107 |
+
if filter_radius > 2:
|
108 |
f0 = signal.medfilt(f0, 3)
|
109 |
elif f0_method == "crepe":
|
110 |
model = "full"
|
|
|
127 |
f0 = torchcrepe.filter.mean(f0, 3)
|
128 |
f0[pd < 0.1] = 0
|
129 |
f0 = f0[0].cpu().numpy()
|
130 |
+
elif f0_method == "rmvpe":
|
131 |
+
if hasattr(self, "model_rmvpe") == False:
|
132 |
+
from rmvpe import RMVPE
|
133 |
+
|
134 |
+
print("loading rmvpe model")
|
135 |
+
self.model_rmvpe = RMVPE(
|
136 |
+
"rmvpe.pt", is_half=self.is_half, device=self.device
|
137 |
+
)
|
138 |
+
f0 = self.model_rmvpe.infer_from_audio(x, thred=0.03)
|
139 |
f0 *= pow(2, f0_up_key / 12)
|
140 |
# with open("test.txt","w")as f:f.write("\n".join([str(i)for i in f0.tolist()]))
|
141 |
tf0 = self.sr // self.window # 每秒f0点数
|
|
|
158 |
) + 1
|
159 |
f0_mel[f0_mel <= 1] = 1
|
160 |
f0_mel[f0_mel > 255] = 255
|
161 |
+
f0_coarse = np.rint(f0_mel).astype(np.int)
|
162 |
return f0_coarse, f0bak # 1-0
|
163 |
|
164 |
def vc(
|
|
|
174 |
big_npy,
|
175 |
index_rate,
|
176 |
version,
|
177 |
+
protect,
|
178 |
): # ,file_index,file_big_npy
|
179 |
feats = torch.from_numpy(audio0)
|
180 |
if self.is_half:
|
|
|
195 |
t0 = ttime()
|
196 |
with torch.no_grad():
|
197 |
logits = model.extract_features(**inputs)
|
198 |
+
feats = model.final_proj(logits[0]) if version == "v1" else logits[0]
|
199 |
+
if protect < 0.5 and pitch != None and pitchf != None:
|
200 |
+
feats0 = feats.clone()
|
201 |
if (
|
202 |
isinstance(index, type(None)) == False
|
203 |
and isinstance(big_npy, type(None)) == False
|
|
|
223 |
)
|
224 |
|
225 |
feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
|
226 |
+
if protect < 0.5 and pitch != None and pitchf != None:
|
227 |
+
feats0 = F.interpolate(feats0.permute(0, 2, 1), scale_factor=2).permute(
|
228 |
+
0, 2, 1
|
229 |
+
)
|
230 |
t1 = ttime()
|
231 |
p_len = audio0.shape[0] // self.window
|
232 |
if feats.shape[1] < p_len:
|
|
|
234 |
if pitch != None and pitchf != None:
|
235 |
pitch = pitch[:, :p_len]
|
236 |
pitchf = pitchf[:, :p_len]
|
237 |
+
|
238 |
+
if protect < 0.5 and pitch != None and pitchf != None:
|
239 |
+
pitchff = pitchf.clone()
|
240 |
+
pitchff[pitchf > 0] = 1
|
241 |
+
pitchff[pitchf < 1] = protect
|
242 |
+
pitchff = pitchff.unsqueeze(-1)
|
243 |
+
feats = feats * pitchff + feats0 * (1 - pitchff)
|
244 |
+
feats = feats.to(feats0.dtype)
|
245 |
p_len = torch.tensor([p_len], device=self.device).long()
|
246 |
with torch.no_grad():
|
247 |
if pitch != None and pitchf != None:
|
|
|
253 |
)
|
254 |
else:
|
255 |
audio1 = (
|
256 |
+
(net_g.infer(feats, p_len, sid)[0][0, 0]).data.cpu().float().numpy()
|
|
|
|
|
|
|
257 |
)
|
258 |
del feats, p_len, padding_mask
|
259 |
if torch.cuda.is_available():
|
|
|
282 |
resample_sr,
|
283 |
rms_mix_rate,
|
284 |
version,
|
285 |
+
protect,
|
286 |
f0_file=None,
|
287 |
):
|
288 |
if (
|
|
|
337 |
sid = torch.tensor(sid, device=self.device).unsqueeze(0).long()
|
338 |
pitch, pitchf = None, None
|
339 |
if if_f0 == 1:
|
340 |
+
pitch, pitchf = self.get_f0(
|
341 |
+
input_audio_path,
|
342 |
+
audio_pad,
|
343 |
+
p_len,
|
344 |
+
f0_up_key,
|
345 |
+
f0_method,
|
346 |
+
filter_radius,
|
347 |
+
inp_f0,
|
348 |
+
)
|
349 |
pitch = pitch[:p_len]
|
350 |
pitchf = pitchf[:p_len]
|
351 |
if self.device == "mps":
|
|
|
370 |
big_npy,
|
371 |
index_rate,
|
372 |
version,
|
373 |
+
protect,
|
374 |
)[self.t_pad_tgt : -self.t_pad_tgt]
|
375 |
)
|
376 |
else:
|
|
|
387 |
big_npy,
|
388 |
index_rate,
|
389 |
version,
|
390 |
+
protect,
|
391 |
)[self.t_pad_tgt : -self.t_pad_tgt]
|
392 |
)
|
393 |
s = t
|
|
|
405 |
big_npy,
|
406 |
index_rate,
|
407 |
version,
|
408 |
+
protect,
|
409 |
)[self.t_pad_tgt : -self.t_pad_tgt]
|
410 |
)
|
411 |
else:
|
|
|
422 |
big_npy,
|
423 |
index_rate,
|
424 |
version,
|
425 |
+
protect,
|
426 |
)[self.t_pad_tgt : -self.t_pad_tgt]
|
427 |
)
|
428 |
audio_opt = np.concatenate(audio_opt)
|
429 |
+
if rms_mix_rate != 1:
|
430 |
+
audio_opt = change_rms(audio, 16000, audio_opt, tgt_sr, rms_mix_rate)
|
431 |
+
if resample_sr >= 16000 and tgt_sr != resample_sr:
|
432 |
audio_opt = librosa.resample(
|
433 |
audio_opt, orig_sr=tgt_sr, target_sr=resample_sr
|
434 |
)
|
435 |
+
audio_max = np.abs(audio_opt).max() / 0.99
|
436 |
+
max_int16 = 32768
|
437 |
+
if audio_max > 1:
|
438 |
+
max_int16 /= audio_max
|
439 |
+
audio_opt = (audio_opt * max_int16).astype(np.int16)
|
440 |
del pitch, pitchf, sid
|
441 |
if torch.cuda.is_available():
|
442 |
torch.cuda.empty_cache()
|