Skip to main content
. 2022 Mar 4;12:703797. doi: 10.3389/fneur.2021.703797

Table 2.

Architecture of the proposed RNN.

Layer Operation Output Parameters #
Input (C ×256) C ×256 ×1
1 LSTM (20) 256 ×20 2,000
Dropout (0.1) 256 x 20 0
2 Time-Distributed Dense (20) 256 x 20 420
3 Global Average Pooling 1D 20 0
4 Dense (1) 1 21

C, Number of channels.