import torch
import torch.nn.functional as F
import torch.nn as nn
from torch.nn import Conv1d, AvgPool1d, Conv2d
from torch.nn.utils import weight_norm, spectral_norm
from .utils import get_padding
LRELU_SLOPE = 0.1
def stft(x, fft_size, hop_size, win_length, window):
"""Perform STFT and convert to magnitude spectrogram.
Args:
x (Tensor): Input signal tensor (B, T).
fft_size (int): FFT size.
hop_size (int): Hop size.
win_length (int): Window length.
window (str): Window function type.
Returns:
Tensor: Magnitude spectrogram (B, #frames, fft_size // 2 + 1).
"""
x_stft = torch.stft(x, fft_size, hop_size, win_length, window, return_complex=True)
real = x_stft[..., 0]
imag = x_stft[..., 1]
return torch.abs(x_stft).transpose(2, 1)
class SpecDiscriminator(nn.Module):
"""docstring for Discriminator."""
def __init__(
self,
fft_size=1024,
shift_size=120,
win_length=600,
window="hann_window",
use_spectral_norm=False,
):
super(SpecDiscriminator, self).__init__()
norm_f = weight_norm if use_spectral_norm == False else spectral_norm
self.fft_size = fft_size
self.shift_size = shift_size
self.win_length = win_length
self.window = getattr(torch, window)(win_length)
self.discriminators = nn.ModuleList(
[
norm_f(nn.Conv2d(1, 32, kernel_size=(3, 9), padding=(1, 4))),
norm_f(
nn.Conv2d(32, 32, kernel_size=(3, 9), stride=(1, 2), padding=(1, 4))
),
norm_f(
nn.Conv2d(32, 32, kernel_size=(3, 9), stride=(1, 2), padding=(1, 4))
),
norm_f(
nn.Conv2d(32, 32, kernel_size=(3, 9), stride=(1, 2), padding=(1, 4))
),
norm_f(
nn.Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
),
]
)
self.out = norm_f(nn.Conv2d(32, 1, 3, 1, 1))
def forward(self, y):
fmap = []
y = y.squeeze(1)
y = stft(
y,
self.fft_size,
self.shift_size,
self.win_length,
self.window.to(y.get_device()),
)
y = y.unsqueeze(1)
for i, d in enumerate(self.discriminators):
y = d(y)
y = F.leaky_relu(y, LRELU_SLOPE)
fmap.append(y)
y = self.out(y)
fmap.append(y)
return torch.flatten(y, 1, -1), fmap
class MultiResSpecDiscriminator(torch.nn.Module):
def __init__(
self,
fft_sizes=[1024, 2048, 512],
hop_sizes=[120, 240, 50],
win_lengths=[600, 1200, 240],
window="hann_window",
):
super(MultiResSpecDiscriminator, self).__init__()
self.discriminators = nn.ModuleList(
[
SpecDiscriminator(fft_sizes[0], hop_sizes[0], win_lengths[0], window),
SpecDiscriminator(fft_sizes[1], hop_sizes[1], win_lengths[1], window),
SpecDiscriminator(fft_sizes[2], hop_sizes[2], win_lengths[2], window),
]
)
def forward(self, y, y_hat):
y_d_rs = []
y_d_gs = []
fmap_rs = []
fmap_gs = []
for i, d in enumerate(self.discriminators):
y_d_r, fmap_r = d(y)
y_d_g, fmap_g = d(y_hat)
y_d_rs.append(y_d_r)
fmap_rs.append(fmap_r)
y_d_gs.append(y_d_g)
fmap_gs.append(fmap_g)
return y_d_rs, y_d_gs, fmap_rs, fmap_gs
class DiscriminatorP(torch.nn.Module):
def __init__(self, period, kernel_size=5, stride=3, use_spectral_norm=False):
super(DiscriminatorP, self).__init__()
self.period = period
norm_f = weight_norm if use_spectral_norm == False else spectral_norm
self.convs = nn.ModuleList(
[
norm_f(
Conv2d(
1,
32,
(kernel_size, 1),
(stride, 1),
padding=(get_padding(5, 1), 0),
)
),
norm_f(
Conv2d(
32,
128,
(kernel_size, 1),
(stride, 1),
padding=(get_padding(5, 1), 0),
)
),
norm_f(
Conv2d(
128,
512,
(kernel_size, 1),
(stride, 1),
padding=(get_padding(5, 1), 0),
)
),
norm_f(
Conv2d(
512,
1024,
(kernel_size, 1),
(stride, 1),
padding=(get_padding(5, 1), 0),
)
),
norm_f(Conv2d(1024, 1024, (kernel_size, 1), 1, padding=(2, 0))),
]
)
self.conv_post = norm_f(Conv2d(1024, 1, (3, 1), 1, padding=(1, 0)))
def forward(self, x):
fmap = []
# 1d to 2d
b, c, t = x.shape
if t % self.period != 0: # pad first
n_pad = self.period - (t % self.period)
x = F.pad(x, (0, n_pad), "reflect")
t = t + n_pad
x = x.view(b, c, t // self.period, self.period)
for l in self.convs:
x = l(x)
x = F.leaky_relu(x, LRELU_SLOPE)
fmap.append(x)
x = self.conv_post(x)
fmap.append(x)
x = torch.flatten(x, 1, -1)
return x, fmap
class MultiPeriodDiscriminator(torch.nn.Module):
def __init__(self):
super(MultiPeriodDiscriminator, self).__init__()
self.discriminators = nn.ModuleList(
[
DiscriminatorP(2),
DiscriminatorP(3),
DiscriminatorP(5),
DiscriminatorP(7),
DiscriminatorP(11),
]
)
def forward(self, y, y_hat):
y_d_rs = []
y_d_gs = []
fmap_rs = []
fmap_gs = []
for i, d in enumerate(self.discriminators):
y_d_r, fmap_r = d(y)
y_d_g, fmap_g = d(y_hat)
y_d_rs.append(y_d_r)
fmap_rs.append(fmap_r)
y_d_gs.append(y_d_g)
fmap_gs.append(fmap_g)
return y_d_rs, y_d_gs, fmap_rs, fmap_gs
class WavLMDiscriminator(nn.Module):
"""docstring for Discriminator."""
def __init__(
self, slm_hidden=768, slm_layers=13, initial_channel=64, use_spectral_norm=False
):
super(WavLMDiscriminator, self).__init__()
norm_f = weight_norm if use_spectral_norm == False else spectral_norm
self.pre = norm_f(
Conv1d(slm_hidden * slm_layers, initial_channel, 1, 1, padding=0)
)
self.convs = nn.ModuleList(
[
norm_f(
nn.Conv1d(
initial_channel, initial_channel * 2, kernel_size=5, padding=2
)
),
norm_f(
nn.Conv1d(
initial_channel * 2,
initial_channel * 4,
kernel_size=5,
padding=2,
)
),
norm_f(
nn.Conv1d(initial_channel * 4, initial_channel * 4, 5, 1, padding=2)
),
]
)
self.conv_post = norm_f(Conv1d(initial_channel * 4, 1, 3, 1, padding=1))
def forward(self, x):
x = self.pre(x)
fmap = []
for l in self.convs:
x = l(x)
x = F.leaky_relu(x, LRELU_SLOPE)
fmap.append(x)
x = self.conv_post(x)
x = torch.flatten(x, 1, -1)
return x