Python hparams.fft_size方法代碼示例

本文整理匯總了Python中hparams.hparams.fft_size方法的典型用法代碼示例。如果您正苦於以下問題：Python hparams.fft_size方法的具體用法？Python hparams.fft_size怎麽用？Python hparams.fft_size使用的例子？那麽, 這裏精選的方法代碼示例或許可以為您提供幫助。您也可以進一步了解該方法所在類hparams.hparams的用法示例。

在下文中一共展示了hparams.fft_size方法的8個代碼示例，這些例子默認根據受歡迎程度排序。您可以為喜歡或者感覺有用的代碼點讚，您的評價將有助於係統推薦出更棒的Python代碼示例。

示例1: _stft

# 需要導入模塊: from hparams import hparams [as 別名]
# 或者: from hparams.hparams import fft_size [as 別名]
def _stft(y):
    return librosa.stft(y=y, n_fft=hparams.fft_size, hop_length=get_hop_size())

開發者ID:candlewill，項目名稱:Griffin_lim，代碼行數:4，代碼來源:audio.py

示例2: _stft

# 需要導入模塊: from hparams import hparams [as 別名]
# 或者: from hparams.hparams import fft_size [as 別名]
def _stft(y):
	return librosa.stft(y=y, n_fft=hparams.fft_size, hop_length=get_hop_size())

開發者ID:rishikksh20，項目名稱:vae_tacotron2，代碼行數:4，代碼來源:audio.py

示例3: _build_mel_basis

# 需要導入模塊: from hparams import hparams [as 別名]
# 或者: from hparams.hparams import fft_size [as 別名]
def _build_mel_basis():
	assert hparams.fmax <= hparams.sample_rate // 2
	return librosa.filters.mel(hparams.sample_rate, hparams.fft_size, n_mels=hparams.num_mels,
							   fmin=hparams.fmin, fmax=hparams.fmax)

開發者ID:rishikksh20，項目名稱:vae_tacotron2，代碼行數:6，代碼來源:audio.py

示例4: _lws_processor

# 需要導入模塊: from hparams import hparams [as 別名]
# 或者: from hparams.hparams import fft_size [as 別名]
def _lws_processor():
    return lws.lws(hparams.fft_size, hparams.hop_size, mode="speech")


# Conversions:

開發者ID:G-Wang，項目名稱:WaveRNN-Pytorch，代碼行數:7，代碼來源:audio.py

示例5: _build_mel_basis

# 需要導入模塊: from hparams import hparams [as 別名]
# 或者: from hparams.hparams import fft_size [as 別名]
def _build_mel_basis():
    if hparams.fmax is not None:
        assert hparams.fmax <= hparams.sample_rate // 2
    return librosa.filters.mel(hparams.sample_rate, hparams.fft_size,
                               fmin=hparams.fmin, fmax=hparams.fmax,
                               n_mels=hparams.num_mels)

開發者ID:G-Wang，項目名稱:WaveRNN-Pytorch，代碼行數:8，代碼來源:audio.py

示例6: _lws_processor

# 需要導入模塊: from hparams import hparams [as 別名]
# 或者: from hparams.hparams import fft_size [as 別名]
def _lws_processor():
    return lws.lws(hparams.fft_size, get_hop_size(), mode="speech")

開發者ID:kastnerkyle，項目名稱:representation_mixing，代碼行數:4，代碼來源:audio.py

示例7: _build_mel_basis

# 需要導入模塊: from hparams import hparams [as 別名]
# 或者: from hparams.hparams import fft_size [as 別名]
def _build_mel_basis():
    assert hparams.fmax <= hparams.sample_rate // 2
    return librosa.filters.mel(hparams.sample_rate, hparams.fft_size,
                               fmin=hparams.fmin, fmax=hparams.fmax,
                               n_mels=hparams.num_mels)

開發者ID:kastnerkyle，項目名稱:representation_mixing，代碼行數:7，代碼來源:audio.py

示例8: _process_utterance

# 需要導入模塊: from hparams import hparams [as 別名]
# 或者: from hparams.hparams import fft_size [as 別名]
def _process_utterance(out_dir, index, wav_path, text):
    # Load the audio to a numpy array:
    wav = audio.load_wav(wav_path)

    if hparams.rescaling:
        wav = wav / np.abs(wav).max() * hparams.rescaling_max

    # Mu-law quantize
    if is_mulaw_quantize(hparams.input_type):
        # [0, quantize_channels)
        out = P.mulaw_quantize(wav, hparams.quantize_channels)

        # Trim silences
        start, end = audio.start_and_end_indices(out, hparams.silence_threshold)
        wav = wav[start:end]
        out = out[start:end]
        constant_values = P.mulaw_quantize(0, hparams.quantize_channels)
        out_dtype = np.int16
    elif is_mulaw(hparams.input_type):
        # [-1, 1]
        out = P.mulaw(wav, hparams.quantize_channels)
        constant_values = P.mulaw(0.0, hparams.quantize_channels)
        out_dtype = np.float32
    else:
        # [-1, 1]
        out = wav
        constant_values = 0.0
        out_dtype = np.float32

    # Compute a mel-scale spectrogram from the trimmed wav:
    # (N, D)
    mel_spectrogram = audio.melspectrogram(wav).astype(np.float32).T
    # lws pads zeros internally before performing stft
    # this is needed to adjust time resolution between audio and mel-spectrogram
    l, r = audio.lws_pad_lr(wav, hparams.fft_size, audio.get_hop_size())

    # zero pad for quantized signal
    out = np.pad(out, (l, r), mode="constant", constant_values=constant_values)
    N = mel_spectrogram.shape[0]
    assert len(out) >= N * audio.get_hop_size()

    # time resolution adjustment
    # ensure length of raw audio is multiple of hop_size so that we can use
    # transposed convolution to upsample
    out = out[:N * audio.get_hop_size()]
    assert len(out) % audio.get_hop_size() == 0

    timesteps = len(out)

    # Write the spectrograms to disk:
    audio_filename = 'ljspeech-audio-%05d.npy' % index
    mel_filename = 'ljspeech-mel-%05d.npy' % index
    np.save(os.path.join(out_dir, audio_filename),
            out.astype(out_dtype), allow_pickle=False)
    np.save(os.path.join(out_dir, mel_filename),
            mel_spectrogram.astype(np.float32), allow_pickle=False)

    # Return a tuple describing this training example:
    return (audio_filename, mel_filename, timesteps, text)

開發者ID:kastnerkyle，項目名稱:representation_mixing，代碼行數:61，代碼來源:ljspeech.py

注：本文中的hparams.hparams.fft_size方法示例由純淨天空整理自Github/MSDocs等開源代碼及文檔管理平台，相關代碼片段篩選自各路編程大神貢獻的開源項目，源碼版權歸原作者所有，傳播和使用請參考對應項目的License；未經允許，請勿轉載。