1
0
mirror of synced 2024-12-25 05:44:57 +01:00
Retrieval-based-Voice-Conve.../infer/lib/uvr5_pack/lib_v5/nets_new.py

134 lines
4.1 KiB
Python
Raw Normal View History

2023-05-28 17:00:29 +02:00
import torch
import torch.nn.functional as F
2023-08-28 09:08:31 +02:00
from torch import nn
2023-06-24 09:26:14 +02:00
from . import layers_new
2023-05-28 17:00:29 +02:00
class BaseNet(nn.Module):
def __init__(
self, nin, nout, nin_lstm, nout_lstm, dilations=((4, 2), (8, 4), (12, 6))
):
2023-05-28 17:00:29 +02:00
super(BaseNet, self).__init__()
2023-06-24 09:26:14 +02:00
self.enc1 = layers_new.Conv2DBNActiv(nin, nout, 3, 1, 1)
self.enc2 = layers_new.Encoder(nout, nout * 2, 3, 2, 1)
self.enc3 = layers_new.Encoder(nout * 2, nout * 4, 3, 2, 1)
self.enc4 = layers_new.Encoder(nout * 4, nout * 6, 3, 2, 1)
self.enc5 = layers_new.Encoder(nout * 6, nout * 8, 3, 2, 1)
2023-05-28 17:00:29 +02:00
2023-06-24 09:26:14 +02:00
self.aspp = layers_new.ASPPModule(nout * 8, nout * 8, dilations, dropout=True)
2023-05-28 17:00:29 +02:00
2023-06-24 09:26:14 +02:00
self.dec4 = layers_new.Decoder(nout * (6 + 8), nout * 6, 3, 1, 1)
self.dec3 = layers_new.Decoder(nout * (4 + 6), nout * 4, 3, 1, 1)
self.dec2 = layers_new.Decoder(nout * (2 + 4), nout * 2, 3, 1, 1)
self.lstm_dec2 = layers_new.LSTMModule(nout * 2, nin_lstm, nout_lstm)
self.dec1 = layers_new.Decoder(nout * (1 + 2) + 1, nout * 1, 3, 1, 1)
2023-05-28 17:00:29 +02:00
def __call__(self, x):
e1 = self.enc1(x)
e2 = self.enc2(e1)
e3 = self.enc3(e2)
e4 = self.enc4(e3)
e5 = self.enc5(e4)
h = self.aspp(e5)
h = self.dec4(h, e4)
h = self.dec3(h, e3)
h = self.dec2(h, e2)
h = torch.cat([h, self.lstm_dec2(h)], dim=1)
h = self.dec1(h, e1)
return h
class CascadedNet(nn.Module):
2023-05-28 17:00:29 +02:00
def __init__(self, n_fft, nout=32, nout_lstm=128):
super(CascadedNet, self).__init__()
self.max_bin = n_fft // 2
self.output_bin = n_fft // 2 + 1
self.nin_lstm = self.max_bin // 2
self.offset = 64
self.stg1_low_band_net = nn.Sequential(
BaseNet(2, nout // 2, self.nin_lstm // 2, nout_lstm),
2023-06-24 09:26:14 +02:00
layers_new.Conv2DBNActiv(nout // 2, nout // 4, 1, 1, 0),
)
self.stg1_high_band_net = BaseNet(
2, nout // 4, self.nin_lstm // 2, nout_lstm // 2
)
2023-05-28 17:00:29 +02:00
self.stg2_low_band_net = nn.Sequential(
BaseNet(nout // 4 + 2, nout, self.nin_lstm // 2, nout_lstm),
2023-06-24 09:26:14 +02:00
layers_new.Conv2DBNActiv(nout, nout // 2, 1, 1, 0),
)
self.stg2_high_band_net = BaseNet(
nout // 4 + 2, nout // 2, self.nin_lstm // 2, nout_lstm // 2
)
2023-05-28 17:00:29 +02:00
self.stg3_full_band_net = BaseNet(
3 * nout // 4 + 2, nout, self.nin_lstm, nout_lstm
)
2023-05-28 17:00:29 +02:00
self.out = nn.Conv2d(nout, 2, 1, bias=False)
self.aux_out = nn.Conv2d(3 * nout // 4, 2, 1, bias=False)
def forward(self, x):
x = x[:, :, : self.max_bin]
2023-05-28 17:00:29 +02:00
bandw = x.size()[2] // 2
l1_in = x[:, :, :bandw]
h1_in = x[:, :, bandw:]
l1 = self.stg1_low_band_net(l1_in)
h1 = self.stg1_high_band_net(h1_in)
aux1 = torch.cat([l1, h1], dim=2)
l2_in = torch.cat([l1_in, l1], dim=1)
h2_in = torch.cat([h1_in, h1], dim=1)
l2 = self.stg2_low_band_net(l2_in)
h2 = self.stg2_high_band_net(h2_in)
aux2 = torch.cat([l2, h2], dim=2)
f3_in = torch.cat([x, aux1, aux2], dim=1)
f3 = self.stg3_full_band_net(f3_in)
mask = torch.sigmoid(self.out(f3))
mask = F.pad(
input=mask,
pad=(0, 0, 0, self.output_bin - mask.size()[2]),
mode="replicate",
2023-05-28 17:00:29 +02:00
)
if self.training:
aux = torch.cat([aux1, aux2], dim=1)
aux = torch.sigmoid(self.aux_out(aux))
aux = F.pad(
input=aux,
pad=(0, 0, 0, self.output_bin - aux.size()[2]),
mode="replicate",
2023-05-28 17:00:29 +02:00
)
return mask, aux
else:
return mask
def predict_mask(self, x):
mask = self.forward(x)
if self.offset > 0:
mask = mask[:, :, :, self.offset : -self.offset]
2023-05-28 17:00:29 +02:00
assert mask.size()[3] > 0
return mask
def predict(self, x, aggressiveness=None):
2023-05-28 17:00:29 +02:00
mask = self.forward(x)
pred_mag = x * mask
if self.offset > 0:
pred_mag = pred_mag[:, :, :, self.offset : -self.offset]
2023-05-28 17:00:29 +02:00
assert pred_mag.size()[3] > 0
return pred_mag