Table 4.

Average accuracies for the benchmark and CNN classifiers trained on imagined vowels.

	Benchmark Methods						CNN Methods
	SVM		RdF		rLDA		Shallow		Deep		EEGNet
	Intra	Inter	Intra	Inter	Intra	Inter	Intra	Inter	Intra	Inter	Intra	Inter
Accuracy	22.23	22.25	23.08	23.23	25.82	26.22	29.62	29.39	29.06	29.58	30.08	30.25
Std.	2.968	3.33	3.88	4.13	3.13	2.32	3.45	2.51	2.58	1.75	2.67	2.73
Max.	26.78	27.12	30.16	29.25	31.68	29.77	35.83	33.36	33.20	32.46	32.38	35.18