Deep Neural Network with Joint Distribution Matching for Cross-Subject Motor Imagery Brain-Computer Interfaces

Table 1

Detailed architecture for the proposed DCJNN.

Layer	Input ()^1	Operations	Output
1
		BatchNorm
		Dropout (0.2)
2
		BatchNorm
		Transpose
		Dropout (0.2)
3
		BatchNorm
		Maxpool2D ()
4
		Maxpool2D ()
5		Flatten
6		Softmax regression	()^2

¹ba denotes the number of samples fed to the network each time. ch denotes the channel. denotes the number of time points. ² stands for the number of classes.