Semisupervised Deep Features of Time-Frequency Maps for Multimodal Emotion Recognition

Table 10

Performance comparison of the proposed method and recently introduced ones.

Authors	Modality	Accuracy
Wu et al. [37]	EEG + EOG	0.866 (two classes)
Hatipoglu Yilmaz and Kose [38]	EEG + EOG	0.915 (two classes)
Ma et al. [39]	EEG + PPS	0.923 (two classes)
Qiu et al. [40]	EEG + PPS	0.856 (two classes)
Li et al. [41]	EEG + PPS	0.949 (two classes)
Zhang et al. [7]	EEG + PPS	0.847 (two classes)
Zhang et al. [8]	EEG + PPS	0.901 (two classes)
Jalal and Peer [12]	PPS	0.842 (four classes)
Cimtay et al. [19]	EEG, GSR, facial	0.915 (four classes)
Proposed method	EEG + PPS	0.953 (two classes)
		0.928 (four classes)