Depression-level assessment from multi-lingual conversational speech data using acoustic and text features

EURASIP Journal on Audio, Speech, and Music Processing

Table 11 Classification results after fusing acoustic- and text-based classifier outputs. Turkish database was used. The results are statistically insignificant except N_min=3 with German

Method	Classes	Precision	Recall	F1-score
*Baseline*	Non-depressed	0.61	0.40	0.48
*MRMR(3)*	Depressed	0.72	0.86	0.78
	Average	0.67	0.63	0.63
*Fusion Tr+Eng*	Non-depressed	0.71	0.63	0.67
N_min=3	Depressed	0.81	0.86	0.83
*ml-MRMR(100)*	Average	0.76	0.75	0.75
*Fusion Tr+Ger*	Non-depressed	0.75	0.55	0.63
N_min=3	Depressed	0.78	0.90	0.84
*ml-MRMR(4)*	Average	0.76	0.73	0.73
*Only Text*	Non-depressed	0.78	0.40	0.53
*Features MRMR(7)*	Depressed	0.74	0.94	0.83
	Average	0.76	0.67	0.68