File size: 19,635 Bytes
ffaf0d2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 |
# import torch
# import torch.nn as nn
# import torch.nn.functional as F
# from audiotools import AudioSignal
# from audiotools import ml
# from audiotools import STFTParams
# from einops import rearrange
# from torch.nn.utils import weight_norm
# def WNConv1d(*args, **kwargs):
# act = kwargs.pop("act", True)
# conv = weight_norm(nn.Conv1d(*args, **kwargs))
# if not act:
# return conv
# return nn.Sequential(conv, nn.LeakyReLU(0.1))
# def WNConv2d(*args, **kwargs):
# act = kwargs.pop("act", True)
# conv = weight_norm(nn.Conv2d(*args, **kwargs))
# if not act:
# return conv
# return nn.Sequential(conv, nn.LeakyReLU(0.1))
# class MPD(nn.Module):
# def __init__(self, period):
# super().__init__()
# self.period = period
# self.convs = nn.ModuleList(
# [
# WNConv2d(1, 32, (5, 1), (3, 1), padding=(2, 0)),
# WNConv2d(32, 128, (5, 1), (3, 1), padding=(2, 0)),
# WNConv2d(128, 512, (5, 1), (3, 1), padding=(2, 0)),
# WNConv2d(512, 1024, (5, 1), (3, 1), padding=(2, 0)),
# WNConv2d(1024, 1024, (5, 1), 1, padding=(2, 0)),
# ]
# )
# self.conv_post = WNConv2d(
# 1024, 1, kernel_size=(3, 1), padding=(1, 0), act=False
# )
# def pad_to_period(self, x):
# t = x.shape[-1]
# x = F.pad(x, (0, self.period - t % self.period), mode="reflect")
# return x
# def forward(self, x):
# fmap = []
# x = self.pad_to_period(x)
# x = rearrange(x, "b c (l p) -> b c l p", p=self.period)
# for layer in self.convs:
# x = layer(x)
# fmap.append(x)
# x = self.conv_post(x)
# fmap.append(x)
# return fmap
# class MSD(nn.Module):
# def __init__(self, rate: int = 1, sample_rate: int = 44100):
# super().__init__()
# self.convs = nn.ModuleList(
# [
# WNConv1d(1, 16, 15, 1, padding=7),
# WNConv1d(16, 64, 41, 4, groups=4, padding=20),
# WNConv1d(64, 256, 41, 4, groups=16, padding=20),
# WNConv1d(256, 1024, 41, 4, groups=64, padding=20),
# WNConv1d(1024, 1024, 41, 4, groups=256, padding=20),
# WNConv1d(1024, 1024, 5, 1, padding=2),
# ]
# )
# self.conv_post = WNConv1d(1024, 1, 3, 1, padding=1, act=False)
# self.sample_rate = sample_rate
# self.rate = rate
# def forward(self, x):
# x = AudioSignal(x, self.sample_rate)
# x.resample(self.sample_rate // self.rate)
# x = x.audio_data
# fmap = []
# for l in self.convs:
# x = l(x)
# fmap.append(x)
# x = self.conv_post(x)
# fmap.append(x)
# return fmap
# BANDS = [(0.0, 0.1), (0.1, 0.25), (0.25, 0.5), (0.5, 0.75), (0.75, 1.0)]
# class MRD(nn.Module):
# def __init__(
# self,
# window_length: int,
# hop_factor: float = 0.25,
# sample_rate: int = 44100,
# bands: list = BANDS,
# ):
# """Complex multi-band spectrogram discriminator.
# Parameters
# ----------
# window_length : int
# Window length of STFT.
# hop_factor : float, optional
# Hop factor of the STFT, defaults to ``0.25 * window_length``.
# sample_rate : int, optional
# Sampling rate of audio in Hz, by default 44100
# bands : list, optional
# Bands to run discriminator over.
# """
# super().__init__()
# self.window_length = window_length
# self.hop_factor = hop_factor
# self.sample_rate = sample_rate
# self.stft_params = STFTParams(
# window_length=window_length,
# hop_length=int(window_length * hop_factor),
# match_stride=True,
# )
# n_fft = window_length // 2 + 1
# bands = [(int(b[0] * n_fft), int(b[1] * n_fft)) for b in bands]
# self.bands = bands
# ch = 32
# convs = lambda: nn.ModuleList(
# [
# WNConv2d(2, ch, (3, 9), (1, 1), padding=(1, 4)),
# WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
# WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
# WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
# WNConv2d(ch, ch, (3, 3), (1, 1), padding=(1, 1)),
# ]
# )
# self.band_convs = nn.ModuleList([convs() for _ in range(len(self.bands))])
# self.conv_post = WNConv2d(ch, 1, (3, 3), (1, 1), padding=(1, 1), act=False)
# def spectrogram(self, x):
# x = AudioSignal(x, self.sample_rate, stft_params=self.stft_params)
# x = torch.view_as_real(x.stft())
# x = rearrange(x, "b 1 f t c -> (b 1) c t f")
# # Split into bands
# x_bands = [x[..., b[0] : b[1]] for b in self.bands]
# return x_bands
# def forward(self, x):
# x_bands = self.spectrogram(x)
# fmap = []
# x = []
# for band, stack in zip(x_bands, self.band_convs):
# for layer in stack:
# band = layer(band)
# fmap.append(band)
# x.append(band)
# x = torch.cat(x, dim=-1)
# x = self.conv_post(x)
# fmap.append(x)
# return fmap
# class Discriminator(ml.BaseModel):
# def __init__(
# self,
# rates: list = [],
# periods: list = [2, 3, 5, 7, 11],
# fft_sizes: list = [2048, 1024, 512],
# sample_rate: int = 44100,
# bands: list = BANDS,
# ):
# """Discriminator that combines multiple discriminators.
# Parameters
# ----------
# rates : list, optional
# sampling rates (in Hz) to run MSD at, by default []
# If empty, MSD is not used.
# periods : list, optional
# periods (of samples) to run MPD at, by default [2, 3, 5, 7, 11]
# fft_sizes : list, optional
# Window sizes of the FFT to run MRD at, by default [2048, 1024, 512]
# sample_rate : int, optional
# Sampling rate of audio in Hz, by default 44100
# bands : list, optional
# Bands to run MRD at, by default `BANDS`
# """
# super().__init__()
# discs = []
# discs += [MPD(p) for p in periods]
# discs += [MSD(r, sample_rate=sample_rate) for r in rates]
# discs += [MRD(f, sample_rate=sample_rate, bands=bands) for f in fft_sizes]
# self.discriminators = nn.ModuleList(discs)
# def preprocess(self, y):
# # Remove DC offset
# y = y - y.mean(dim=-1, keepdims=True)
# # Peak normalize the volume of input audio
# y = 0.8 * y / (y.abs().max(dim=-1, keepdim=True)[0] + 1e-9)
# return y
# def forward(self, x):
# x = self.preprocess(x)
# fmaps = [d(x) for d in self.discriminators]
# return fmaps
# if __name__ == "__main__":
# disc = Discriminator()
# x = torch.zeros(1, 1, 44100)
# results = disc(x)
# for i, result in enumerate(results):
# print(f"disc{i}")
# for i, r in enumerate(result):
# print(r.shape, r.mean(), r.min(), r.max())
# print()
import torch
import torch.nn as nn
import torch.nn.functional as F
from audiotools import AudioSignal, STFTParams
from audiotools import ml
from einops import rearrange
from torch.nn.utils import weight_norm
import torchaudio
import nnAudio.features as features
from munch import Munch
BANDS = [(0.0, 0.1), (0.1, 0.25), (0.25, 0.5), (0.5, 0.75), (0.75, 1.0)]
def WNConv1d(*args, **kwargs):
act = kwargs.pop("act", True)
conv = weight_norm(nn.Conv1d(*args, **kwargs))
if not act:
return conv
return nn.Sequential(conv, nn.LeakyReLU(0.1))
def WNConv2d(*args, **kwargs):
act = kwargs.pop("act", True)
conv = weight_norm(nn.Conv2d(*args, **kwargs))
if not act:
return conv
return nn.Sequential(conv, nn.LeakyReLU(0.1))
def get_padding(kernel_size, dilation=1):
return int((kernel_size * dilation - dilation) / 2)
def get_2d_padding(kernel_size, dilation=(1, 1)):
return (int((kernel_size[0] * dilation[0] - dilation[0]) / 2),
int((kernel_size[1] * dilation[1] - dilation[1]) / 2))
class NormConv2d(nn.Module):
"""Conv2d with normalization"""
def __init__(self, in_channels, out_channels, kernel_size, stride=1,
padding=0, dilation=1, groups=1, bias=True, norm="weight_norm"):
super().__init__()
self.conv = nn.Conv2d(in_channels, out_channels, kernel_size,
stride, padding, dilation, groups, bias)
if norm == "weight_norm":
self.conv = weight_norm(self.conv)
def forward(self, x):
return self.conv(x)
class MPD(nn.Module):
def __init__(self, period):
super().__init__()
self.period = period
self.convs = nn.ModuleList([
WNConv2d(1, 32, (5, 1), (3, 1), padding=(2, 0)),
WNConv2d(32, 128, (5, 1), (3, 1), padding=(2, 0)),
WNConv2d(128, 512, (5, 1), (3, 1), padding=(2, 0)),
WNConv2d(512, 1024, (5, 1), (3, 1), padding=(2, 0)),
WNConv2d(1024, 1024, (5, 1), 1, padding=(2, 0)),
])
self.conv_post = WNConv2d(1024, 1, kernel_size=(3, 1), padding=(1, 0), act=False)
def pad_to_period(self, x):
t = x.shape[-1]
x = F.pad(x, (0, self.period - t % self.period), mode="reflect")
return x
def forward(self, x):
fmap = []
x = self.pad_to_period(x)
x = rearrange(x, "b c (l p) -> b c l p", p=self.period)
for layer in self.convs:
x = layer(x)
fmap.append(x)
x = self.conv_post(x)
fmap.append(x)
return fmap
class MSD(nn.Module):
def __init__(self, rate: int = 1, sample_rate: int = 44100):
super().__init__()
self.convs = nn.ModuleList([
WNConv1d(1, 16, 15, 1, padding=7),
WNConv1d(16, 64, 41, 4, groups=4, padding=20),
WNConv1d(64, 256, 41, 4, groups=16, padding=20),
WNConv1d(256, 1024, 41, 4, groups=64, padding=20),
WNConv1d(1024, 1024, 41, 4, groups=256, padding=20),
WNConv1d(1024, 1024, 5, 1, padding=2),
])
self.conv_post = WNConv1d(1024, 1, 3, 1, padding=1, act=False)
self.sample_rate = sample_rate
self.rate = rate
def forward(self, x):
x = AudioSignal(x, self.sample_rate)
x.resample(self.sample_rate // self.rate)
x = x.audio_data
fmap = []
for l in self.convs:
x = l(x)
fmap.append(x)
x = self.conv_post(x)
fmap.append(x)
return fmap
class DiscriminatorCQT(nn.Module):
def __init__(self, cfg, hop_length, n_octaves, bins_per_octave):
super().__init__()
self.cfg = cfg
self.filters = cfg.filters
self.max_filters = cfg.max_filters
self.filters_scale = cfg.filters_scale
self.kernel_size = (3, 9)
self.dilations = cfg.dilations
self.stride = (1, 2)
self.in_channels = cfg.in_channels
self.out_channels = cfg.out_channels
self.fs = cfg.sampling_rate
self.hop_length = hop_length
self.n_octaves = n_octaves
self.bins_per_octave = bins_per_octave
self.cqt_transform = features.cqt.CQT2010v2(
sr=self.fs * 2,
hop_length=self.hop_length,
n_bins=self.bins_per_octave * self.n_octaves,
bins_per_octave=self.bins_per_octave,
output_format="Complex",
pad_mode="constant",
)
self.conv_pres = nn.ModuleList()
for i in range(self.n_octaves):
self.conv_pres.append(
NormConv2d(
self.in_channels * 2, # Real + Imaginary
self.in_channels * 2,
kernel_size=self.kernel_size,
padding=get_2d_padding(self.kernel_size),
norm="weight_norm",
)
)
self.convs = nn.ModuleList()
self.convs.append(
NormConv2d(
self.in_channels * 2,
self.filters,
kernel_size=self.kernel_size,
padding=get_2d_padding(self.kernel_size),
)
)
in_chs = min(self.filters_scale * self.filters, self.max_filters)
for i, dilation in enumerate(self.dilations):
out_chs = min((self.filters_scale ** (i + 1)) * self.filters, self.max_filters)
self.convs.append(
NormConv2d(
in_chs,
out_chs,
kernel_size=self.kernel_size,
stride=self.stride,
dilation=(dilation, 1),
padding=get_2d_padding(self.kernel_size, (dilation, 1)),
norm="weight_norm",
)
)
in_chs = out_chs
out_chs = min(
(self.filters_scale ** (len(self.dilations) + 1)) * self.filters,
self.max_filters,
)
self.convs.append(
NormConv2d(
in_chs,
out_chs,
kernel_size=(self.kernel_size[0], self.kernel_size[0]),
padding=get_2d_padding((self.kernel_size[0], self.kernel_size[0])),
norm="weight_norm",
)
)
self.conv_post = NormConv2d(
out_chs,
self.out_channels,
kernel_size=(self.kernel_size[0], self.kernel_size[0]),
padding=get_2d_padding((self.kernel_size[0], self.kernel_size[0])),
norm="weight_norm",
)
self.activation = torch.nn.LeakyReLU(negative_slope=0.1)
self.resample = torchaudio.transforms.Resample(
orig_freq=self.fs, new_freq=self.fs * 2
)
def forward(self, x):
fmap = []
x = self.resample(x)
z = self.cqt_transform(x)
z_amplitude = z[:, :, :, 0].unsqueeze(1)
z_phase = z[:, :, :, 1].unsqueeze(1)
z = torch.cat([z_amplitude, z_phase], dim=1)
z = rearrange(z, "b c w t -> b c t w")
latent_z = []
for i in range(self.n_octaves):
octave_band = z[:, :, :, i * self.bins_per_octave : (i + 1) * self.bins_per_octave]
processed_band = self.conv_pres[i](octave_band)
latent_z.append(processed_band)
latent_z = torch.cat(latent_z, dim=-1)
for i, l in enumerate(self.convs):
latent_z = l(latent_z)
latent_z = self.activation(latent_z)
fmap.append(latent_z)
latent_z = self.conv_post(latent_z)
fmap.append(latent_z)
return fmap
class MultiScaleSubbandCQT(nn.Module):
"""CQT discriminator at multiple scales"""
def __init__(self, sample_rate=44100):
super().__init__()
cfg = Munch({
"hop_lengths": [1024, 512, 512],
"sampling_rate": sample_rate,
"filters": 32,
"max_filters": 1024,
"filters_scale": 1,
"dilations": [1, 2, 4],
"in_channels": 1,
"out_channels": 1,
"n_octaves": [10, 10, 10],
"bins_per_octaves": [24, 36, 48],
})
self.cfg = cfg
self.discriminators = nn.ModuleList([
DiscriminatorCQT(
cfg,
hop_length=cfg.hop_lengths[i],
n_octaves=cfg.n_octaves[i],
bins_per_octave=cfg.bins_per_octaves[i],
)
for i in range(len(cfg.hop_lengths))
])
def forward(self, x):
fmap = []
for disc in self.discriminators:
fmap.extend(disc(x))
return fmap
BANDS = [(0.0, 0.1), (0.1, 0.25), (0.25, 0.5), (0.5, 0.75), (0.75, 1.0)]
class MRD(nn.Module):
def __init__(self, window_length: int, hop_factor: float = 0.25,
sample_rate: int = 44100, bands: list = BANDS):
"""Multi-resolution spectrogram discriminator."""
super().__init__()
self.window_length = window_length
self.hop_factor = hop_factor
self.sample_rate = sample_rate
self.stft_params = STFTParams(
window_length=window_length,
hop_length=int(window_length * hop_factor),
match_stride=True,
)
n_fft = window_length // 2 + 1
bands = [(int(b[0] * n_fft), int(b[1] * n_fft)) for b in bands]
self.bands = bands
ch = 32
convs = lambda: nn.ModuleList([
WNConv2d(2, ch, (3, 9), (1, 1), padding=(1, 4)),
WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
WNConv2d(ch, ch, (3, 9), (1, 2), padding=(1, 4)),
WNConv2d(ch, ch, (3, 3), (1, 1), padding=(1, 1)),
])
self.band_convs = nn.ModuleList([convs() for _ in range(len(self.bands))])
self.conv_post = WNConv2d(ch, 1, (3, 3), (1, 1), padding=(1, 1), act=False)
def spectrogram(self, x):
x = AudioSignal(x, self.sample_rate, stft_params=self.stft_params)
x = torch.view_as_real(x.stft())
x = rearrange(x, "b 1 f t c -> (b 1) c t f")
x_bands = [x[..., b[0] : b[1]] for b in self.bands]
return x_bands
def forward(self, x):
x_bands = self.spectrogram(x)
fmap = []
x = []
for band, stack in zip(x_bands, self.band_convs):
for layer in stack:
band = layer(band)
fmap.append(band)
x.append(band)
x = torch.cat(x, dim=-1)
x = self.conv_post(x)
fmap.append(x)
return fmap
class Discriminator(ml.BaseModel):
def __init__(
self,
rates: list = [],
periods: list = [2, 3, 5, 7, 11],
fft_sizes: list = [2048, 1024, 512],
sample_rate: int = 44100,
):
"""Discriminator combining MPD, MSD, MRD and CQT.
Parameters
----------
rates : list, optional
Sampling rates for MSD, by default []
periods : list, optional
Periods for MPD, by default [2, 3, 5, 7, 11]
fft_sizes : list, optional
FFT sizes for MRD, by default [2048, 1024, 512]
sample_rate : int, optional
Sampling rate of audio in Hz, by default 44100
"""
super().__init__()
discs = []
# Time-domain discriminators
discs += [MPD(p) for p in periods]
discs += [MSD(r, sample_rate=sample_rate) for r in rates]
# Frequency-domain discriminators (both STFT and CQT)
discs += [MRD(f, sample_rate=sample_rate) for f in fft_sizes]
discs += [MultiScaleSubbandCQT(sample_rate=sample_rate)]
self.discriminators = nn.ModuleList(discs)
def preprocess(self, y):
# Remove DC offset
y = y - y.mean(dim=-1, keepdims=True)
# Peak normalize
y = 0.8 * y / (y.abs().max(dim=-1, keepdim=True)[0] + 1e-9)
return y
def forward(self, x):
x = self.preprocess(x)
fmaps = [d(x) for d in self.discriminators]
return fmaps |