espnet.nets.pytorch_backend.tacotron2.decoder.Postnet

source

class espnet.nets.pytorch_backend.tacotron2.decoder.Postnet(idim, odim, n_layers=5, n_chans=512, n_filts=5, dropout_rate=0.5, use_batch_norm=True)

Bases: Module

Postnet module for Spectrogram prediction network.

This is a module of Postnet in Spectrogram prediction network, which described in Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions. The Postnet predicts refines the predicted Mel-filterbank of the decoder, which helps to compensate the detail structure of spectrogram.

Initialize postnet module.

• Parameters:
  • idim (int) – Dimension of the inputs.
  • odim (int) – Dimension of the outputs.
  • n_layers (int , optional) – The number of layers.
  • n_filts (int , optional) – The number of filter size.
  • n_units (int , optional) – The number of filter channels.
  • use_batch_norm (bool , optional) – Whether to use batch normalization..
  • dropout_rate (float , optional) – Dropout rate..

forward(xs)

Calculate forward propagation.

• Parameters:xs (Tensor) – Batch of the sequences of padded input tensors (B, idim, Tmax).
• Returns: Batch of padded output tensor. (B, odim, Tmax).
• Return type: Tensor