espnet.nets.pytorch_backend.frontends.dnn_beamformer.DNN_Beamformer

source

class espnet.nets.pytorch_backend.frontends.dnn_beamformer.DNN_Beamformer(bidim, btype='blstmp', blayers=3, bunits=300, bprojs=320, bnmask=2, dropout_rate=0.0, badim=320, ref_channel: int = -1, beamformer_type='mvdr')

Bases: Module

DNN mask based Beamformer

Citation: : Multichannel End-to-end Speech Recognition; T. Ochiai et al., 2017; https://arxiv.org/abs/1703.04783

Initialize internal Module state, shared by both nn.Module and ScriptModule.

forward(data: ComplexTensor, ilens: LongTensor) → Tuple[ComplexTensor, LongTensor, ComplexTensor]

The forward function

Notation: : B: Batch C: Channel T: Time or Sequence length F: Freq

• Parameters:
  • data (ComplexTensor) – (B, T, C, F)
  • ilens (torch.Tensor) – (B,)
• Returns: (B, T, F) ilens (torch.Tensor): (B,)
• Return type: enhanced (ComplexTensor)