Python tf.signal.inverse_mdct用法及代碼示例

計算 mdcts 的逆修改 DCT。

用法

tf.signal.inverse_mdct(
    mdcts, window_fn=tf.signal.vorbis_window, norm=None, name=None
)

要重建原始波形，應將相同的窗口函數與 mdct 和 inverse_mdct 一起使用。

示例用法：

@tf.function
def compare_round_trip():
  samples = 1000
  frame_length = 400
  halflen = frame_length // 2
  waveform = tf.random.normal(dtype=tf.float32, shape=[samples])
  waveform_pad = tf.pad(waveform, [[halflen, 0],])
  mdct = tf.signal.mdct(waveform_pad, frame_length, pad_end=True,
                        window_fn=tf.signal.vorbis_window)
  inverse_mdct = tf.signal.inverse_mdct(mdct,
                                        window_fn=tf.signal.vorbis_window)
  inverse_mdct = inverse_mdct[halflen:halflen + samples]
  return waveform, inverse_mdct
waveform, inverse_mdct = compare_round_trip()
np.allclose(waveform.numpy(), inverse_mdct.numpy(), rtol=1e-3, atol=1e-4)
True

使用 TPU/GPU-compatible 操作實現並支持漸變。