Multichannel Speech Enhancement in Vehicle Environment Based on Interchannel Attention Mechanism

Figure 4

The structure of the BiLSTM block. (a) The module contains BiLSTM, linear, and GroupNorm layers, where the numbers in parentheses indicate the size of the output dimension. (b) The internal structure diagram of the BiLSTM layer, where A refers to the LSTM module, is the input data, is the output of the hidden layer, and is the output result BiLSTM.