File size: 5,759 Bytes
6127b48 |
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 |
import librosa
import matplotlib
import matplotlib.pyplot as plt
import numpy as np
import torch
matplotlib.use("Agg")
def plot_alignment(alignment, info=None, fig_size=(16, 10), title=None, output_fig=False):
if isinstance(alignment, torch.Tensor):
alignment_ = alignment.detach().cpu().numpy().squeeze()
else:
alignment_ = alignment
alignment_ = alignment_.astype(np.float32) if alignment_.dtype == np.float16 else alignment_
fig, ax = plt.subplots(figsize=fig_size)
im = ax.imshow(alignment_.T, aspect="auto", origin="lower", interpolation="none")
fig.colorbar(im, ax=ax)
xlabel = "Decoder timestep"
if info is not None:
xlabel += "\n\n" + info
plt.xlabel(xlabel)
plt.ylabel("Encoder timestep")
# plt.yticks(range(len(text)), list(text))
plt.tight_layout()
if title is not None:
plt.title(title)
if not output_fig:
plt.close()
return fig
def plot_spectrogram(spectrogram, ap=None, fig_size=(16, 10), output_fig=False):
if isinstance(spectrogram, torch.Tensor):
spectrogram_ = spectrogram.detach().cpu().numpy().squeeze().T
else:
spectrogram_ = spectrogram.T
spectrogram_ = spectrogram_.astype(np.float32) if spectrogram_.dtype == np.float16 else spectrogram_
if ap is not None:
spectrogram_ = ap.denormalize(spectrogram_) # pylint: disable=protected-access
fig = plt.figure(figsize=fig_size)
plt.imshow(spectrogram_, aspect="auto", origin="lower")
plt.colorbar()
plt.tight_layout()
if not output_fig:
plt.close()
return fig
def plot_pitch(pitch, spectrogram, ap=None, fig_size=(30, 10), output_fig=False):
"""Plot pitch curves on top of the spectrogram.
Args:
pitch (np.array): Pitch values.
spectrogram (np.array): Spectrogram values.
Shapes:
pitch: :math:`(T,)`
spec: :math:`(C, T)`
"""
if isinstance(spectrogram, torch.Tensor):
spectrogram_ = spectrogram.detach().cpu().numpy().squeeze().T
else:
spectrogram_ = spectrogram.T
spectrogram_ = spectrogram_.astype(np.float32) if spectrogram_.dtype == np.float16 else spectrogram_
if ap is not None:
spectrogram_ = ap.denormalize(spectrogram_) # pylint: disable=protected-access
old_fig_size = plt.rcParams["figure.figsize"]
if fig_size is not None:
plt.rcParams["figure.figsize"] = fig_size
fig, ax = plt.subplots()
ax.imshow(spectrogram_, aspect="auto", origin="lower")
ax.set_xlabel("time")
ax.set_ylabel("spec_freq")
ax2 = ax.twinx()
ax2.plot(pitch, linewidth=5.0, color="red")
ax2.set_ylabel("F0")
plt.rcParams["figure.figsize"] = old_fig_size
if not output_fig:
plt.close()
return fig
def plot_avg_pitch(pitch, chars, fig_size=(30, 10), output_fig=False):
"""Plot pitch curves on top of the input characters.
Args:
pitch (np.array): Pitch values.
chars (str): Characters to place to the x-axis.
Shapes:
pitch: :math:`(T,)`
"""
old_fig_size = plt.rcParams["figure.figsize"]
if fig_size is not None:
plt.rcParams["figure.figsize"] = fig_size
fig, ax = plt.subplots()
x = np.array(range(len(chars)))
my_xticks = chars
plt.xticks(x, my_xticks)
ax.set_xlabel("characters")
ax.set_ylabel("freq")
ax2 = ax.twinx()
ax2.plot(pitch, linewidth=5.0, color="red")
ax2.set_ylabel("F0")
plt.rcParams["figure.figsize"] = old_fig_size
if not output_fig:
plt.close()
return fig
def visualize(
alignment,
postnet_output,
text,
hop_length,
CONFIG,
tokenizer,
stop_tokens=None,
decoder_output=None,
output_path=None,
figsize=(8, 24),
output_fig=False,
):
"""Intended to be used in Notebooks."""
if decoder_output is not None:
num_plot = 4
else:
num_plot = 3
label_fontsize = 16
fig = plt.figure(figsize=figsize)
plt.subplot(num_plot, 1, 1)
plt.imshow(alignment.T, aspect="auto", origin="lower", interpolation=None)
plt.xlabel("Decoder timestamp", fontsize=label_fontsize)
plt.ylabel("Encoder timestamp", fontsize=label_fontsize)
# compute phoneme representation and back
if CONFIG.use_phonemes:
seq = tokenizer.text_to_ids(text)
text = tokenizer.ids_to_text(seq)
print(text)
plt.yticks(range(len(text)), list(text))
plt.colorbar()
if stop_tokens is not None:
# plot stopnet predictions
plt.subplot(num_plot, 1, 2)
plt.plot(range(len(stop_tokens)), list(stop_tokens))
# plot postnet spectrogram
plt.subplot(num_plot, 1, 3)
librosa.display.specshow(
postnet_output.T,
sr=CONFIG.audio["sample_rate"],
hop_length=hop_length,
x_axis="time",
y_axis="linear",
fmin=CONFIG.audio["mel_fmin"],
fmax=CONFIG.audio["mel_fmax"],
)
plt.xlabel("Time", fontsize=label_fontsize)
plt.ylabel("Hz", fontsize=label_fontsize)
plt.tight_layout()
plt.colorbar()
if decoder_output is not None:
plt.subplot(num_plot, 1, 4)
librosa.display.specshow(
decoder_output.T,
sr=CONFIG.audio["sample_rate"],
hop_length=hop_length,
x_axis="time",
y_axis="linear",
fmin=CONFIG.audio["mel_fmin"],
fmax=CONFIG.audio["mel_fmax"],
)
plt.xlabel("Time", fontsize=label_fontsize)
plt.ylabel("Hz", fontsize=label_fontsize)
plt.tight_layout()
plt.colorbar()
if output_path:
print(output_path)
fig.savefig(output_path)
plt.close()
if not output_fig:
plt.close()
|