|
import argparse |
|
|
|
import parselmouth |
|
import soundfile |
|
import librosa |
|
import torch |
|
import numpy as np |
|
|
|
import cluster |
|
import utils |
|
import json |
|
|
|
from inference.infer_tool import Svc |
|
import logging |
|
logging.getLogger('numba').setLevel(logging.WARNING) |
|
|
|
sample_rate = 44100 |
|
hop_len = 512 |
|
|
|
|
|
def getc(audio_data, hmodel, spk): |
|
devive = torch.device("cuda" if torch.cuda.is_available() else "cpu") |
|
wav, sr = audio_data |
|
wav = librosa.resample(wav,sr, 16000) |
|
wav = torch.from_numpy(wav).unsqueeze(0).to(devive) |
|
c = utils.get_vec_units(hmodel, wav).cpu().squeeze(0) |
|
c = utils.repeat_expand_2d(c, int((wav.shape[1] * sample_rate / 16000) // hop_len)).numpy() |
|
c = cluster.get_cluster_center_result(c.transpose(), spk) |
|
return c |
|
|
|
def get_f0(audio_data, p_len=None, f0_up_key=0): |
|
x, sr = audio_data |
|
x = librosa.resample(x,sr, sample_rate) |
|
if p_len is None: |
|
p_len = x.shape[0] // hop_len |
|
else: |
|
assert abs(p_len - x.shape[0] // hop_len) < 3, (p_len, x.shape) |
|
time_step = hop_len / sample_rate * 1000 |
|
f0_min = 50 |
|
f0_max = 1100 |
|
f0_mel_min = 1127 * np.log(1 + f0_min / 700) |
|
f0_mel_max = 1127 * np.log(1 + f0_max / 700) |
|
|
|
f0 = parselmouth.Sound(x, sample_rate).to_pitch_ac( |
|
time_step=time_step / 1000, voicing_threshold=0.6, |
|
pitch_floor=f0_min, pitch_ceiling=f0_max).selected_array['frequency'] |
|
|
|
pad_size = (p_len - len(f0) + 1) // 2 |
|
if (pad_size > 0 or p_len - len(f0) - pad_size > 0): |
|
f0 = np.pad(f0, [[pad_size, p_len - len(f0) - pad_size]], mode='constant') |
|
|
|
f0 *= pow(2, f0_up_key / 12) |
|
f0_mel = 1127 * np.log(1 + f0 / 700) |
|
f0_mel[f0_mel > 0] = (f0_mel[f0_mel > 0] - f0_mel_min) * 254 / (f0_mel_max - f0_mel_min) + 1 |
|
f0_mel[f0_mel <= 1] = 1 |
|
f0_mel[f0_mel > 255] = 255 |
|
f0_coarse = np.rint(f0_mel).astype(np.int) |
|
return f0_coarse, f0 |
|
|
|
hmodel = utils.load_vec_model() |
|
|
|
|
|
def infer(spk, raw, model, trans=0): |
|
audio_data = librosa.load(raw, sr=None) |
|
c = getc(audio_data, hmodel, spk) |
|
_, f0 = get_f0(audio_data, c.shape[0], trans) |
|
|
|
c = torch.from_numpy(c).float().unsqueeze(0).transpose(1, 2) |
|
f0 = torch.from_numpy(f0).float().unsqueeze(0) |
|
res = model.infer(spk, c, f0) |
|
|
|
return res |
|
|
|
|
|
if __name__ == '__main__': |
|
step = 34000 |
|
model_path = f"logs/G_{step}.pth" |
|
config_path = "configs/config.json" |
|
svc_model = Svc(model_path, config_path) |
|
raw = 'raw/君の知らない物語.wav' |
|
spk = "taffy" |
|
|
|
infer(spk, raw, svc_model) |
|
|
