Skip to content

D
DeepSpeech

PaddlePaddle / DeepSpeech
大约 2 年 前同步成功

通知 210
Star 8425
Fork 1598
D
DeepSpeech
提交 63b44947 编写于 作者: Y YangZhou
浏览文件
操作

fix optional bind, add sox_effects

上级 c37782c1
隐藏空白更改
内联 并排
Showing with 450 addition and 23 deletion
paddlespeech/audio/backends/sox_io_backend.py
浏览文件 @ 63b44947
......@@ -8,8 +8,7 @@ from paddle import Tensor
from .common import AudioMetaData
from paddlespeech.audio._internal import module_utils as _mod_utils
from paddlespeech.audio._paddleaudio import get_info_file
from paddlespeech.audio._paddleaudio import get_info_fileobj
from paddlespeech.aduio import _paddleaudio as paddleaudio
#https://github.com/pytorch/audio/blob/main/torchaudio/backend/sox_io_backend.py
......@@ -43,26 +42,38 @@ _fallback_load_filebj = _fail_load_fileobj
@_mod_utils.requires_sox()
def load(
filepath: Union[str, Path],
out: Optional[Tensor]=None,
normalization: Union[bool, float, Callable]=True,
channels_first: bool=True,
num_frames: int=0,
offset: int=0,
filetype: Optional[str]=None, ) -> Tuple[Tensor, int]:
raise RuntimeError("No audio I/O backend is available.")
filepath: str,
frame_offset: int = 0,
num_frames: int=-1,
normalize: bool = True,
channels_first: bool = True,
format: Optional[str]=None, ) -> Tuple[Tensor, int]:
ret = paddleaudio.sox_io_load_audio_file(
filepath, frame_offset, num_frames, normalize, channels_first, format
)
if ret is not None:
return ret
return _fallback_load(filepath, frame_offset, num_frames, normalize, channels_first, format)
@_mod_utils.requires_sox()
def save(filepath: str,
src: Tensor,
sample_rate: int,
precision: int = 16,
channels_first: bool = True) -> None:
raise RuntimeError("No audio I/O backend is available.")
frame_offset: int = 0,
num_frames: int = -1,
normalize: bool = True,
channels_first: bool = True,
format: Optional[str] = None) -> Tuple[Tensor, int]:
ret = paddleaudio.sox_io_load_audio_file(
filepath, frame_offset, num_frames, normalize, channels_first, format
)
if ret is not None:
return ret
return _fallback_load(filepath, frame_offset, num_frames, normalize, channels_first, format)
@_mod_utils.requires_sox()
def info(filepath: str, format: Optional[str]) -> None:
sinfo = paddleaudio._paddleaudio.get_info_file(filepath, format)
sinfo = paddleaudio.get_info_file(filepath, format)
if sinfo is not None:
return AudioMetaData(*sinfo)
return _fallback_info(filepath, format)
paddlespeech/audio/sox_effects/__init__.py 0 → 100644
浏览文件 @ 63b44947
from paddlespeech.audio._internal import module_utils as _mod_utils
from .sox_effects import (
apply_effects_file,
apply_effects_tensor,
effect_names,
init_sox_effects,
shutdown_sox_effects,
)
if _mod_utils.is_sox_available():
import atexit
init_sox_effects()
atexit.register(shutdown_sox_effects)
__all__ = [
"init_sox_effects",
"shutdown_sox_effects",
"effect_names",
"apply_effects_tensor",
"apply_effects_file",
]
paddlespeech/audio/sox_effects/sox_effects.py 0 → 100644
浏览文件 @ 63b44947
import os
from typing import List, Optional, Tuple
from paddlespeech.audio._internal import module_utils as _mod_utils
from paddlespeech.audio.utils.sox_utils import list_effects
from paddlespeech.audio import _paddleaudio as paddleaudio
#code is from: https://github.com/pytorch/audio/blob/main/torchaudio/sox_effects/sox_effects.py
@_mod_utils.requires_sox()
def init_sox_effects():
"""Initialize resources required to use sox effects.
Note:
You do not need to call this function manually. It is called automatically.
Once initialized, you do not need to call this function again across the multiple uses of
sox effects though it is safe to do so as long as :func:`shutdown_sox_effects` is not called yet.
Once :func:`shutdown_sox_effects` is called, you can no longer use SoX effects and initializing
again will result in error.
"""
paddleaudio.sox_effects_initialize_sox_effects()
@_mod_utils.requires_sox()
def shutdown_sox_effects():
"""Clean up resources required to use sox effects.
Note:
You do not need to call this function manually. It is called automatically.
It is safe to call this function multiple times.
Once :py:func:`shutdown_sox_effects` is called, you can no longer use SoX effects and
initializing again will result in error.
"""
paddleaudio.sox_effects_shutdown_sox_effects()
@_mod_utils.requires_sox()
def effect_names() -> List[str]:
"""Gets list of valid sox effect names
Returns:
List[str]: list of available effect names.
Example
>>> paddleaudio.sox_effects.effect_names()
['allpass', 'band', 'bandpass', ... ]
"""
return list(list_effects().keys())
@_mod_utils.requires_sox()
def apply_effects_tensor(
tensor: torch.Tensor,
sample_rate: int,
effects: List[List[str]],
channels_first: bool = True,
) -> Tuple[torch.Tensor, int]:
"""Apply sox effects to given Tensor
.. devices:: CPU
.. properties:: TorchScript
Note:
This function only works on CPU Tensors.
This function works in the way very similar to ``sox`` command, however there are slight
differences. For example, ``sox`` command adds certain effects automatically (such as
``rate`` effect after ``speed`` and ``pitch`` and other effects), but this function does
only applies the given effects. (Therefore, to actually apply ``speed`` effect, you also
need to give ``rate`` effect with desired sampling rate.).
Args:
tensor (torch.Tensor): Input 2D CPU Tensor.
sample_rate (int): Sample rate
effects (List[List[str]]): List of effects.
channels_first (bool, optional): Indicates if the input Tensor's dimension is
`[channels, time]` or `[time, channels]`
Returns:
(Tensor, int): Resulting Tensor and sample rate.
The resulting Tensor has the same ``dtype`` as the input Tensor, and
the same channels order. The shape of the Tensor can be different based on the
effects applied. Sample rate can also be different based on the effects applied.
Example - Basic usage
>>>
>>> # Defines the effects to apply
>>> effects = [
... ['gain', '-n'], # normalises to 0dB
... ['pitch', '5'], # 5 cent pitch shift
... ['rate', '8000'], # resample to 8000 Hz
... ]
>>>
>>> # Generate pseudo wave:
>>> # normalized, channels first, 2ch, sampling rate 16000, 1 second
>>> sample_rate = 16000
>>> waveform = 2 * torch.rand([2, sample_rate * 1]) - 1
>>> waveform.shape
torch.Size([2, 16000])
>>> waveform
tensor([[ 0.3138, 0.7620, -0.9019, ..., -0.7495, -0.4935, 0.5442],
[-0.0832, 0.0061, 0.8233, ..., -0.5176, -0.9140, -0.2434]])
>>>
>>> # Apply effects
>>> waveform, sample_rate = apply_effects_tensor(
... wave_form, sample_rate, effects, channels_first=True)
>>>
>>> # Check the result
>>> # The new waveform is sampling rate 8000, 1 second.
>>> # normalization and channel order are preserved
>>> waveform.shape
torch.Size([2, 8000])
>>> waveform
tensor([[ 0.5054, -0.5518, -0.4800, ..., -0.0076, 0.0096, -0.0110],
[ 0.1331, 0.0436, -0.3783, ..., -0.0035, 0.0012, 0.0008]])
>>> sample_rate
8000
Example - Torchscript-able transform
>>>
>>> # Use `apply_effects_tensor` in `torch.nn.Module` and dump it to file,
>>> # then run sox effect via Torchscript runtime.
>>>
>>> class SoxEffectTransform(torch.nn.Module):
... effects: List[List[str]]
...
... def __init__(self, effects: List[List[str]]):
... super().__init__()
... self.effects = effects
...
... def forward(self, tensor: torch.Tensor, sample_rate: int):
... return sox_effects.apply_effects_tensor(
... tensor, sample_rate, self.effects)
...
...
>>> # Create transform object
>>> effects = [
... ["lowpass", "-1", "300"], # apply single-pole lowpass filter
... ["rate", "8000"], # change sample rate to 8000
... ]
>>> transform = SoxEffectTensorTransform(effects, input_sample_rate)
>>>
>>> # Dump it to file and load
>>> path = 'sox_effect.zip'
>>> torch.jit.script(trans).save(path)
>>> transform = torch.jit.load(path)
>>>
>>>> # Run transform
>>> waveform, input_sample_rate = paddleaudio.load("input.wav")
>>> waveform, sample_rate = transform(waveform, input_sample_rate)
>>> assert sample_rate == 8000
"""
return paddleaudio.sox_effects_apply_effects_tensor(tensor, sample_rate, effects, channels_first)
@_mod_utils.requires_sox()
def apply_effects_file(
path: str,
effects: List[List[str]],
normalize: bool = True,
channels_first: bool = True,
format: Optional[str] = None,
) -> Tuple[torch.Tensor, int]:
"""Apply sox effects to the audio file and load the resulting data as Tensor
.. devices:: CPU
.. properties:: TorchScript
Note:
This function works in the way very similar to ``sox`` command, however there are slight
differences. For example, ``sox`` commnad adds certain effects automatically (such as
``rate`` effect after ``speed``, ``pitch`` etc), but this function only applies the given
effects. Therefore, to actually apply ``speed`` effect, you also need to give ``rate``
effect with desired sampling rate, because internally, ``speed`` effects only alter sampling
rate and leave samples untouched.
Args:
path (path-like object or file-like object):
Source of audio data. When the function is not compiled by TorchScript,
(e.g. ``torch.jit.script``), the following types are accepted:
* ``path-like``: file path
* ``file-like``: Object with ``read(size: int) -> bytes`` method,
which returns byte string of at most ``size`` length.
When the function is compiled by TorchScript, only ``str`` type is allowed.
Note: This argument is intentionally annotated as ``str`` only for
TorchScript compiler compatibility.
effects (List[List[str]]): List of effects.
normalize (bool, optional):
When ``True``, this function always return ``float32``, and sample values are
normalized to ``[-1.0, 1.0]``.
If input file is integer WAV, giving ``False`` will change the resulting Tensor type to
integer type. This argument has no effect for formats other
than integer WAV type.
channels_first (bool, optional): When True, the returned Tensor has dimension `[channel, time]`.
Otherwise, the returned Tensor's dimension is `[time, channel]`.
format (str or None, optional):
Override the format detection with the given format.
Providing the argument might help when libsox can not infer the format
from header or extension,
Returns:
(Tensor, int): Resulting Tensor and sample rate.
If ``normalize=True``, the resulting Tensor is always ``float32`` type.
If ``normalize=False`` and the input audio file is of integer WAV file, then the
resulting Tensor has corresponding integer type. (Note 24 bit integer type is not supported)
If ``channels_first=True``, the resulting Tensor has dimension `[channel, time]`,
otherwise `[time, channel]`.
Example - Basic usage
>>>
>>> # Defines the effects to apply
>>> effects = [
... ['gain', '-n'], # normalises to 0dB
... ['pitch', '5'], # 5 cent pitch shift
... ['rate', '8000'], # resample to 8000 Hz
... ]
>>>
>>> # Apply effects and load data with channels_first=True
>>> waveform, sample_rate = apply_effects_file("data.wav", effects, channels_first=True)
>>>
>>> # Check the result
>>> waveform.shape
torch.Size([2, 8000])
>>> waveform
tensor([[ 5.1151e-03, 1.8073e-02, 2.2188e-02, ..., 1.0431e-07,
-1.4761e-07, 1.8114e-07],
[-2.6924e-03, 2.1860e-03, 1.0650e-02, ..., 6.4122e-07,
-5.6159e-07, 4.8103e-07]])
>>> sample_rate
8000
Example - Apply random speed perturbation to dataset
>>>
>>> # Load data from file, apply random speed perturbation
>>> class RandomPerturbationFile(torch.utils.data.Dataset):
... \"\"\"Given flist, apply random speed perturbation
...
... Suppose all the input files are at least one second long.
... \"\"\"
... def __init__(self, flist: List[str], sample_rate: int):
... super().__init__()
... self.flist = flist
... self.sample_rate = sample_rate
...
... def __getitem__(self, index):
... speed = 0.5 + 1.5 * random.randn()
... effects = [
... ['gain', '-n', '-10'], # apply 10 db attenuation
... ['remix', '-'], # merge all the channels
... ['speed', f'{speed:.5f}'], # duration is now 0.5 ~ 2.0 seconds.
... ['rate', f'{self.sample_rate}'],
... ['pad', '0', '1.5'], # add 1.5 seconds silence at the end
... ['trim', '0', '2'], # get the first 2 seconds
... ]
... waveform, _ = paddleaudio.sox_effects.apply_effects_file(
... self.flist[index], effects)
... return waveform
...
... def __len__(self):
... return len(self.flist)
...
>>> dataset = RandomPerturbationFile(file_list, sample_rate=8000)
>>> loader = torch.utils.data.DataLoader(dataset, batch_size=32)
>>> for batch in loader:
>>> pass
"""
if not torch.jit.is_scripting():
if hasattr(path, "read"):
ret = paddleaudio._paddleaudio.apply_effects_fileobj(path, effects, normalize, channels_first, format)
if ret is None:
raise RuntimeError("Failed to load audio from {}".format(path))
return ret
path = os.fspath(path)
ret = paddleaudio.sox_effects_apply_effects_file(path, effects, normalize, channels_first, format)
if ret is not None:
return ret
raise RuntimeError("Failed to load audio from {}".format(path))
\ No newline at end of file
paddlespeech/audio/src/pybind/pybind.cpp
浏览文件 @ 63b44947
......@@ -5,17 +5,23 @@
#include "paddlespeech/audio/src/pybind/sox/io.h"
#include "paddlespeech/audio/src/pybind/sox/effects.h"
#include "paddlespeech/audio/third_party/kaldi/feat/feature-fbank.h"
#include <pybind11/stl.h>
#include <pybind11/complex.h>
#incldue <pybind11/functional.h>
#include <pybind11/chrono.h>
#include <pybind11/pybind11.h>
// `tl::optional`
namespace pybind11 { namespace detail {
template <typename T>
struct type_caster<tl::optional<T>> : optional_caster<tl::optional<T>> {};
}}
PYBIND11_MODULE(_paddleaudio, m) {
#ifdef INCLUDE_SOX
m.def("get_info_file",
&paddleaudio::sox_io::get_info_file,
"Get metadata of audio file.");
m.def("get_info_fileobj",
// support obj later
/*m.def("get_info_fileobj",
&paddleaudio::sox_io::get_info_fileobj,
"Get metadata of audio in file object.");
m.def("load_audio_fileobj",
......@@ -24,6 +30,7 @@ PYBIND11_MODULE(_paddleaudio, m) {
m.def("save_audio_fileobj",
&paddleaudio::sox_io::save_audio_fileobj,
"Save audio to file obj.");
*/
// sox io
m.def("sox_io_get_info", &paddleaudio::sox_io::get_info_file);
m.def(
......@@ -58,9 +65,9 @@ PYBIND11_MODULE(_paddleaudio, m) {
&paddleaudio::sox_utils::get_buffer_size);
// effect
m.def("apply_effects_fileobj",
&paddleaudio::sox_effects::apply_effects_fileobj,
"Decode audio data from file-like obj and apply effects.");
//m.def("apply_effects_fileobj",
// &paddleaudio::sox_effects::apply_effects_fileobj,
// "Decode audio data from file-like obj and apply effects.");
m.def("sox_effects_initialize_sox_effects",
&paddleaudio::sox_effects::initialize_sox_effects);
m.def(
......
paddlespeech/audio/utils/sox_utils.py 0 → 100644
浏览文件 @ 63b44947
from typing import Dict, List
from paddlespeech.audio._internal import module_utils as _mod_utils
from paddlespeech.audio import _paddleaudio
@_mod_utils.requires_sox()
def set_seed(seed: int):
"""Set libsox's PRNG
Args:
seed (int): seed value. valid range is int32.
See Also:
http://sox.sourceforge.net/sox.html
"""
_paddleaudio.sox_utils_set_seed(seed)
@_mod_utils.requires_sox()
def set_verbosity(verbosity: int):
"""Set libsox's verbosity
Args:
verbosity (int): Set verbosity level of libsox.
* ``1`` failure messages
* ``2`` warnings
* ``3`` details of processing
* ``4``-``6`` increasing levels of debug messages
See Also:
http://sox.sourceforge.net/sox.html
"""
_paddleaudio.sox_utils_set_verbosity(verbosity)
@_mod_utils.requires_sox()
def set_buffer_size(buffer_size: int):
"""Set buffer size for sox effect chain
Args:
buffer_size (int): Set the size in bytes of the buffers used for processing audio.
See Also:
http://sox.sourceforge.net/sox.html
"""
_paddleaudio.sox_utils_set_buffer_size(buffer_size)
@_mod_utils.requires_sox()
def set_use_threads(use_threads: bool):
"""Set multithread option for sox effect chain
Args:
use_threads (bool): When ``True``, enables ``libsox``'s parallel effects channels processing.
To use mutlithread, the underlying ``libsox`` has to be compiled with OpenMP support.
See Also:
http://sox.sourceforge.net/sox.html
"""
_paddleaudio.sox_utils_set_use_threads(use_threads)
@_mod_utils.requires_sox()
def list_effects() -> Dict[str, str]:
"""List the available sox effect names
Returns:
Dict[str, str]: Mapping from ``effect name`` to ``usage``
"""
return dict(_paddleaudio.sox_utils_list_effects())
@_mod_utils.requires_sox()
def list_read_formats() -> List[str]:
"""List the supported audio formats for read
Returns:
List[str]: List of supported audio formats
"""
return _paddleaudio.sox_utils_list_read_formats()
@_mod_utils.requires_sox()
def list_write_formats() -> List[str]:
"""List the supported audio formats for write
Returns:
List[str]: List of supported audio formats
"""
return _paddleaudio.sox_utils_list_write_formats()
@_mod_utils.requires_sox()
def get_buffer_size() -> int:
"""Get buffer size for sox effect chain
Returns:
int: size in bytes of buffers used for processing audio.
"""
return _paddleaudio.sox_utils_get_buffer_size()
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或