nvidia.dali.fn.preemphasis_filter — NVIDIA DALI (original) (raw)

nvidia.dali.fn.preemphasis_filter(__input, /, *, border='clamp', bytes_per_sample_hint=[0], dtype=DALIDataType.FLOAT, preemph_coeff=0.97, preserve=False, device=None, name=None)#

Applies preemphasis filter to the input data.

This filter, in simple form, can be expressed by the formula:

Y[t] = X[t] - coeff * X[t-1] if t > 1 Y[t] = X[t] - coeff * X_border if t == 0

with X and Y being the input and output signal, respectively.

The value of X_border depends on the border argument:

X_border = 0 if border_type == 'zero' X_border = X[0] if border_type == 'clamp' X_border = X[1] if border_type == 'reflect'

Supported backends

• ‘cpu’
• ‘gpu’

Parameters:

__input¶ (TensorList) – Input to the operator.

Keyword Arguments:

• border¶ (str, optional, default = ‘clamp’) – Border value policy. Possible values are "zero", "clamp", "reflect".
• bytes_per_sample_hint¶ (int or list of int, optional, default = [0]) –
  Output size hint, in bytes per sample.
  If specified, the operator’s outputs residing in GPU or page-locked host memory will be preallocated to accommodate a batch of samples of this size.
• dtype¶ (nvidia.dali.types.DALIDataType, optional, default = DALIDataType.FLOAT) – Data type for the output.
• preemph_coeff¶ (float or TensorList of float, optional, default = 0.97) – Preemphasis coefficient coeff.
• preserve¶ (bool, optional, default = False) – Prevents the operator from being removed from the graph even if its outputs are not used.