Attention-Based DSC-ConvLSTM for Multiclass Motor Imagery Classification

Table 3

Within-subject classification accuracy_bcic iv 2a.

Author	Algorithm	Accuracy (%)
		A01	A02	A03	A04	A05	A06	A07	A08	A09	Mean
Sakhavi S	FBCSP	76.11	44.33	81.52	66.30	58.96	50.75	85.69	77.35	75.63	68.0
R. T. Schirr	Deep ConvNet	69.75	45.44	87.46	63.85	52.04	53.39	87.46	83.46	83.99	69.6
	Shallow ConvNet	81.60	45.49	88.54	68.06	60.42	51.74	88.54	81.25	79.51	71.7
V. J. Lawhern	EEGNet-4, 2	80.56	48.96	87.50	64.93	62.50	58.68	87.85	77.78	81.25	72.2
	EEGNet-8, 2	85.76	43.06	92.36	62.50	63.19	62.85	83.33	73.96	76.39	72.5
Ours	Ours	91.32	44.10	90.97	67.71	60.07	56.25	89.58	82.99	79.51	73.7