dequantize

class diffsptk.InverseUniformQuantization(abs_max: float = 1, n_bit: int = 8, quantizer: str | int = 'mid-rise')

See this page for details.

Parameters:

abs_maxfloat > 0

The absolute maximum value of the original waveform.

n_bitint >= 1

The number of quantization bits.

quantizer[‘mid-rise’, ‘mid-tread’]

The quantizer type.

forward(y: Tensor) → Tensor

Dequantize the input waveform.

Parameters:

yTensor [shape=(…,)]

The quantized waveform.

Returns:

outTensor [shape=(…,)]

The dequantized waveform.

Examples

>>> x = diffsptk.ramp(-4, 4)
>>> x
tensor([-4., -3., -2., -1.,  0.,  1.,  2.,  3.,  4.])
>>> quantize = diffsptk.UniformQuantization(4, 2)
>>> dequantize = diffsptk.InverseUniformQuantization(4, 2)
>>> x2 = dequantize(quantize(x))
>>> x2
tensor([-3., -3., -1., -1.,  1.,  1.,  3.,  3.,  3.])

diffsptk.functional.dequantize(y: Tensor, abs_max: float = 1, n_bit: int = 8, quantizer: str = 'mid-rise') → Tensor

Dequantize the input waveform.

Parameters:

yTensor [shape=(…,)]

The quantized waveform.

abs_maxfloat > 0

The absolute maximum value of the original waveform.

n_bitint >= 1

The number of quantization bits.

quantizer[‘mid-rise’, ‘mid-tread’]

The quantizer type.

Returns:

outTensor [shape=(…,)]

The dequantized waveform.

See also

iulaw quantize