VOICEVN / main /inference /preprocess.py
AnhP's picture
Upload 65 files
98bb602 verified
raw
history blame
14.7 kB
import os
import sys
import time
import torch
import logging
import librosa
import argparse
import logging.handlers
import numpy as np
import soundfile as sf
import multiprocessing
import noisereduce as nr
from tqdm import tqdm
from scipy import signal
from scipy.io import wavfile
from distutils.util import strtobool
from concurrent.futures import ProcessPoolExecutor, as_completed
now_directory = os.getcwd()
sys.path.append(now_directory)
from main.configs.config import Config
logger = logging.getLogger(__name__)
logging.getLogger("numba.core.byteflow").setLevel(logging.ERROR)
logging.getLogger("numba.core.ssa").setLevel(logging.ERROR)
logging.getLogger("numba.core.interpreter").setLevel(logging.ERROR)
OVERLAP = 0.3
MAX_AMPLITUDE = 0.9
ALPHA = 0.75
HIGH_PASS_CUTOFF = 48
SAMPLE_RATE_16K = 16000
config = Config()
per = 3.0 if config.is_half else 3.7
translations = config.translations
def parse_arguments() -> tuple:
parser = argparse.ArgumentParser()
parser.add_argument("--model_name", type=str, required=True)
parser.add_argument("--dataset_path", type=str, default="./dataset")
parser.add_argument("--sample_rate", type=int, required=True)
parser.add_argument("--cpu_cores", type=int, default=2)
parser.add_argument("--cut_preprocess", type=lambda x: bool(strtobool(x)), default=True)
parser.add_argument("--process_effects", type=lambda x: bool(strtobool(x)), default=False)
parser.add_argument("--clean_dataset", type=lambda x: bool(strtobool(x)), default=False)
parser.add_argument("--clean_strength", type=float, default=0.7)
args = parser.parse_args()
return args
def load_audio(file, sample_rate):
try:
file = file.strip(" ").strip('"').strip("\n").strip('"').strip(" ")
audio, sr = sf.read(file)
if len(audio.shape) > 1: audio = librosa.to_mono(audio.T)
if sr != sample_rate: audio = librosa.resample(audio, orig_sr=sr, target_sr=sample_rate)
except Exception as e:
raise RuntimeError(f"{translations['errors_loading_audio']}: {e}")
return audio.flatten()
class Slicer:
def __init__(self, sr, threshold = -40.0, min_length = 5000, min_interval = 300, hop_size = 20, max_sil_kept = 5000):
if not min_length >= min_interval >= hop_size: raise ValueError(translations["min_length>=min_interval>=hop_size"])
if not max_sil_kept >= hop_size: raise ValueError(translations["max_sil_kept>=hop_size"])
min_interval = sr * min_interval / 1000
self.threshold = 10 ** (threshold / 20.0)
self.hop_size = round(sr * hop_size / 1000)
self.win_size = min(round(min_interval), 4 * self.hop_size)
self.min_length = round(sr * min_length / 1000 / self.hop_size)
self.min_interval = round(min_interval / self.hop_size)
self.max_sil_kept = round(sr * max_sil_kept / 1000 / self.hop_size)
def _apply_slice(self, waveform, begin, end):
start_idx = begin * self.hop_size
if len(waveform.shape) > 1:
end_idx = min(waveform.shape[1], end * self.hop_size)
return waveform[:, start_idx:end_idx]
else:
end_idx = min(waveform.shape[0], end * self.hop_size)
return waveform[start_idx:end_idx]
def slice(self, waveform):
samples = waveform.mean(axis=0) if len(waveform.shape) > 1 else waveform
if samples.shape[0] <= self.min_length: return [waveform]
rms_list = get_rms(y=samples, frame_length=self.win_size, hop_length=self.hop_size).squeeze(0)
sil_tags = []
silence_start, clip_start = None, 0
for i, rms in enumerate(rms_list):
if rms < self.threshold:
if silence_start is None: silence_start = i
continue
if silence_start is None: continue
is_leading_silence = silence_start == 0 and i > self.max_sil_kept
need_slice_middle = (i - silence_start >= self.min_interval and i - clip_start >= self.min_length)
if not is_leading_silence and not need_slice_middle:
silence_start = None
continue
if i - silence_start <= self.max_sil_kept:
pos = rms_list[silence_start : i + 1].argmin() + silence_start
if silence_start == 0: sil_tags.append((0, pos))
else: sil_tags.append((pos, pos))
clip_start = pos
elif i - silence_start <= self.max_sil_kept * 2:
pos = rms_list[i - self.max_sil_kept : silence_start + self.max_sil_kept + 1].argmin()
pos += i - self.max_sil_kept
pos_l = (rms_list[silence_start : silence_start + self.max_sil_kept + 1].argmin() + silence_start)
pos_r = (rms_list[i - self.max_sil_kept : i + 1].argmin() + i - self.max_sil_kept)
if silence_start == 0:
sil_tags.append((0, pos_r))
clip_start = pos_r
else:
sil_tags.append((min(pos_l, pos), max(pos_r, pos)))
clip_start = max(pos_r, pos)
else:
pos_l = (rms_list[silence_start : silence_start + self.max_sil_kept + 1].argmin() + silence_start)
pos_r = (rms_list[i - self.max_sil_kept : i + 1].argmin() + i - self.max_sil_kept)
if silence_start == 0: sil_tags.append((0, pos_r))
else: sil_tags.append((pos_l, pos_r))
clip_start = pos_r
silence_start = None
total_frames = rms_list.shape[0]
if (silence_start is not None and total_frames - silence_start >= self.min_interval):
silence_end = min(total_frames, silence_start + self.max_sil_kept)
pos = rms_list[silence_start : silence_end + 1].argmin() + silence_start
sil_tags.append((pos, total_frames + 1))
if not sil_tags: return [waveform]
else:
chunks = []
if sil_tags[0][0] > 0: chunks.append(self._apply_slice(waveform, 0, sil_tags[0][0]))
for i in range(len(sil_tags) - 1):
chunks.append(self._apply_slice(waveform, sil_tags[i][1], sil_tags[i + 1][0]))
if sil_tags[-1][1] < total_frames: chunks.append(self._apply_slice(waveform, sil_tags[-1][1], total_frames))
return chunks
def get_rms(y, frame_length=2048, hop_length=512, pad_mode="constant"):
padding = (int(frame_length // 2), int(frame_length // 2))
y = np.pad(y, padding, mode=pad_mode)
axis = -1
out_strides = y.strides + tuple([y.strides[axis]])
x_shape_trimmed = list(y.shape)
x_shape_trimmed[axis] -= frame_length - 1
out_shape = tuple(x_shape_trimmed) + tuple([frame_length])
xw = np.lib.stride_tricks.as_strided(y, shape=out_shape, strides=out_strides)
target_axis = axis - 1 if axis < 0 else axis + 1
xw = np.moveaxis(xw, -1, target_axis)
slices = [slice(None)] * xw.ndim
slices[axis] = slice(0, None, hop_length)
x = xw[tuple(slices)]
power = np.mean(np.abs(x) ** 2, axis=-2, keepdims=True)
return np.sqrt(power)
class PreProcess:
def __init__(self, sr, exp_dir, per):
self.slicer = Slicer(sr=sr, threshold=-42, min_length=1500, min_interval=400, hop_size=15, max_sil_kept=500)
self.sr = sr
self.b_high, self.a_high = signal.butter(N=5, Wn=HIGH_PASS_CUTOFF, btype="high", fs=self.sr)
self.per = per
self.exp_dir = exp_dir
self.device = "cpu"
self.gt_wavs_dir = os.path.join(exp_dir, "sliced_audios")
self.wavs16k_dir = os.path.join(exp_dir, "sliced_audios_16k")
os.makedirs(self.gt_wavs_dir, exist_ok=True)
os.makedirs(self.wavs16k_dir, exist_ok=True)
def _normalize_audio(self, audio: torch.Tensor):
tmp_max = torch.abs(audio).max()
if tmp_max > 2.5: return None
return (audio / tmp_max * (MAX_AMPLITUDE * ALPHA)) + (1 - ALPHA) * audio
def process_audio_segment(self, normalized_audio: np.ndarray, sid, idx0, idx1):
if normalized_audio is None:
logs(f"{sid}-{idx0}-{idx1}-filtered")
return
wavfile.write(os.path.join(self.gt_wavs_dir, f"{sid}_{idx0}_{idx1}.wav"), self.sr, normalized_audio.astype(np.float32))
audio_16k = librosa.resample(normalized_audio, orig_sr=self.sr, target_sr=SAMPLE_RATE_16K)
wavfile.write(os.path.join(self.wavs16k_dir, f"{sid}_{idx0}_{idx1}.wav"), SAMPLE_RATE_16K, audio_16k.astype(np.float32))
def process_audio(self, path, idx0, sid, cut_preprocess, process_effects, clean_dataset, clean_strength):
try:
audio = load_audio(path, self.sr)
if process_effects:
audio = signal.lfilter(self.b_high, self.a_high, audio)
audio = self._normalize_audio(audio)
if clean_dataset: audio = nr.reduce_noise(y=audio, sr=self.sr, prop_decrease=clean_strength)
idx1 = 0
if cut_preprocess:
for audio_segment in self.slicer.slice(audio):
i = 0
while 1:
start = int(self.sr * (self.per - OVERLAP) * i)
i += 1
if len(audio_segment[start:]) > (self.per + OVERLAP) * self.sr:
tmp_audio = audio_segment[start : start + int(self.per * self.sr)]
self.process_audio_segment(tmp_audio, sid, idx0, idx1)
idx1 += 1
else:
tmp_audio = audio_segment[start:]
self.process_audio_segment(tmp_audio, sid, idx0, idx1)
idx1 += 1
break
else: self.process_audio_segment(audio, sid, idx0, idx1)
except Exception as e:
raise RuntimeError(f"{translations['process_audio_error']}: {e}")
def process_file(args):
pp, file, cut_preprocess, process_effects, clean_dataset, clean_strength = (args)
file_path, idx0, sid = file
pp.process_audio(file_path, idx0, sid, cut_preprocess, process_effects, clean_dataset, clean_strength)
def preprocess_training_set(input_root, sr, num_processes, exp_dir, per, cut_preprocess, process_effects, clean_dataset, clean_strength):
start_time = time.time()
pp = PreProcess(sr, exp_dir, per)
logger.info(translations["start_preprocess"].format(num_processes=num_processes))
files = []
idx = 0
for root, _, filenames in os.walk(input_root):
try:
sid = 0 if root == input_root else int(os.path.basename(root))
for f in filenames:
if f.lower().endswith((".wav", ".mp3", ".flac", ".ogg")):
files.append((os.path.join(root, f), idx, sid))
idx += 1
except ValueError:
raise ValueError(f"{translations['not_integer']} '{os.path.basename(root)}'.")
with tqdm(total=len(files), desc=translations["preprocess"]) as pbar:
with ProcessPoolExecutor(max_workers=num_processes) as executor:
futures = [
executor.submit(
process_file,
(
pp,
file,
cut_preprocess,
process_effects,
clean_dataset,
clean_strength,
),
)
for file in files
]
for future in as_completed(futures):
try:
future.result()
except Exception as e:
raise RuntimeError(f"{translations['process_error']}: {e}")
pbar.update(1)
elapsed_time = time.time() - start_time
logger.info(translations["preprocess_success"].format(elapsed_time=f"{elapsed_time:.2f}"))
if __name__ == "__main__":
args = parse_arguments()
experiment_directory = os.path.join("assets", "logs", args.model_name)
num_processes = args.cpu_cores
num_processes = multiprocessing.cpu_count() if num_processes is None else int(num_processes)
dataset = args.dataset_path
sample_rate = args.sample_rate
cut_preprocess = args.cut_preprocess
preprocess_effects = args.process_effects
clean_dataset = args.clean_dataset
clean_strength = args.clean_strength
os.makedirs(experiment_directory, exist_ok=True)
if len([f for f in os.listdir(os.path.join(dataset)) if os.path.isfile(os.path.join(dataset, f)) and f.lower().endswith((".wav", ".mp3", ".flac", ".ogg"))]) < 1: raise FileNotFoundError("Không tìm thấy dữ liệu")
log_file = os.path.join(experiment_directory, "preprocess.log")
if logger.hasHandlers(): logger.handlers.clear()
else:
console_handler = logging.StreamHandler()
console_formatter = logging.Formatter(fmt="\n%(asctime)s.%(msecs)03d | %(levelname)s | %(module)s | %(message)s", datefmt="%Y-%m-%d %H:%M:%S")
console_handler.setFormatter(console_formatter)
console_handler.setLevel(logging.INFO)
file_handler = logging.handlers.RotatingFileHandler(log_file, maxBytes=5*1024*1024, backupCount=3, encoding='utf-8')
file_formatter = logging.Formatter(fmt="\n%(asctime)s.%(msecs)03d | %(levelname)s | %(module)s | %(message)s", datefmt="%Y-%m-%d %H:%M:%S")
file_handler.setFormatter(file_formatter)
file_handler.setLevel(logging.DEBUG)
logger.addHandler(console_handler)
logger.addHandler(file_handler)
logger.setLevel(logging.DEBUG)
logger.debug(f"{translations['modelname']}: {args.model_name}")
logger.debug(f"{translations['export_process']}: {experiment_directory}")
logger.debug(f"{translations['dataset_folder']}: {dataset}")
logger.debug(f"{translations['pretrain_sr']}: {sample_rate}")
logger.debug(f"{translations['cpu_core']}: {num_processes}")
logger.debug(f"{translations['split_audio']}: {cut_preprocess}")
logger.debug(f"{translations['preprocess_effect']}: {preprocess_effects}")
logger.debug(f"{translations['clear_audio']}: {clean_dataset}")
if clean_dataset: logger.debug(f"{translations['clean_strength']}: {clean_strength}")
try:
preprocess_training_set(dataset, sample_rate, num_processes, experiment_directory, per, cut_preprocess, preprocess_effects, clean_dataset, clean_strength)
except Exception as e:
logger.error(f"{translations['process_audio_error']} {e}")
logger.info(f"{translations['preprocess_model_success']} {args.model_name}")