Recognition of Emotions in Mexican Spanish Speech: An Approach Based on Acoustic Modelling of Emotion-Specific Vowels

Figure 4

MFCCs obtained for some emotion-specific vowels of a male speaker from the built emotional speech database. Description of identifiers: _e = anger, _f = happiness, _n = neutral, and _t = sadness. HTK coding parameters: TARGETKIND = MFCC_E_D_A, WINDOWSIZE = 250000.0, USEHAMMING = T, PREEMCOEF = 0.97, NUMCHANS = 26, CEPLIFTER = 22, and NUMCEPS = 12.

(a)

(b)

(c)

(d)

(e)