Spaces:
Running
Running
File size: 4,238 Bytes
16f0ad7 |
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 |
# *****************************************************************************
# Copyright (c) 2018, NVIDIA CORPORATION. All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are met:
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# * Neither the name of the NVIDIA CORPORATION nor the
# names of its contributors may be used to endorse or promote products
# derived from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
# WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
# DISCLAIMED. IN NO EVENT SHALL NVIDIA CORPORATION BE LIABLE FOR ANY
# DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
# (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
# LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
# ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
# SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
#
# *****************************************************************************\
import torch
import random
import common.layers as layers
from common.utils import load_wav_to_torch, load_filepaths_and_text, to_gpu
class MelAudioLoader(torch.utils.data.Dataset):
"""
1) loads audio,text pairs
2) computes mel-spectrograms from audio files.
"""
def __init__(self,
dataset_path,
audiopaths_and_text,
segment_length,
n_mel_channels,
max_wav_value,
sampling_rate,
filter_length,
hop_length,
win_length,
mel_fmin,
mel_fmax,
args):
self.audiopaths_and_text = load_filepaths_and_text(dataset_path, audiopaths_and_text)
self.max_wav_value = max_wav_value
self.sampling_rate = sampling_rate
self.stft = layers.TacotronSTFT(
filter_length, hop_length, win_length,
n_mel_channels, sampling_rate, mel_fmin,
mel_fmax)
self.segment_length = segment_length
random.seed(1234)
random.shuffle(self.audiopaths_and_text)
def get_mel_audio_pair(self, filename):
audio, sampling_rate = load_wav_to_torch(filename)
if sampling_rate != self.stft.sampling_rate:
raise ValueError("{} {} SR doesn't match target {} SR".format(
sampling_rate, self.stft.sampling_rate))
# Take segment
if audio.size(0) >= self.segment_length:
max_audio_start = audio.size(0) - self.segment_length
audio_start = random.randint(0, max_audio_start)
audio = audio[audio_start:audio_start+self.segment_length]
else:
audio = torch.nn.functional.pad(
audio, (0, self.segment_length - audio.size(0)), 'constant').data
audio = audio / self.max_wav_value
audio_norm = audio.unsqueeze(0)
audio_norm = torch.autograd.Variable(audio_norm, requires_grad=False)
melspec = self.stft.mel_spectrogram(audio_norm)
melspec = melspec.squeeze(0)
return (melspec, audio, len(audio))
def __getitem__(self, index):
return self.get_mel_audio_pair(self.audiopaths_and_text[index][0])
def __len__(self):
return len(self.audiopaths_and_text)
def batch_to_gpu(batch):
x, y, len_y = batch
x = to_gpu(x).float()
y = to_gpu(y).float()
len_y = to_gpu(torch.sum(len_y))
return ((x, y), y, len_y)
|