Evaluation of influence of spectral and prosodic features on GMM classification of Czech and Slovak emotional speech

EURASIP Journal on Audio, Speech, and Music Processing

Table 15 Influence of used type of the feature set on the gender recognition error rate; summarized for all emotions

Error rate (%)/feature set	P1	P2	P3	P4	P5	P6	P7
Minimum	0	2.56	0	0	0	0	0
Maximum	38.86	53.57	40.56	32.14	69.74	74.62	40.14
Mean	16.77	18.72	18.69	13.45	29.84	32.19	18.29