Identification of Visual Imagery by Electroencephalography Based on Empirical Mode Decomposition and an Autoregressive Model

Table 5

VI task (paradigm), feature-extraction method, classification method, and classification accuracy in VI-BCI research.

Author	VI tasks	Feature-extraction method	Classification method	Classification accuracy
Kosmyna et al.	Flower; hammer	Power spectrum	SpecCSP	52%
Neuper et al.	Visualizing the movement of one’s hand; resting state	Frequency band	DSLVQ	56%
Koizumi et al.	UAV moves in three planes (up/down, left/right, front/back)	PSD	SVM	84.6%
Sousa et al.	Static point; dynamic point moving vertically in two directions; and dynamic point moving vertically in four directions	Power-spectrum energy	SVM	87.64%
This research	Static star and star moving right	HHT, AR model, and EMD + AR	SVM	HHT: 68.14  3.06%; AR: 56.29  2.73%; EMD + AR: 78.40  2.07%