Skip to main content

Table 7 The performance comparisons between different frequency warping methods used in the proposed GMM-HMM/DNN-HMM EASR system for the Persian ESD and CREMA-D datasets and the system of Sheikhan et al. [15] for various acoustic features and emotional states

From: Feature compensation based on the normalization of vocal tract length for the improvement of emotion-affected speech recognition

Sheikhan et al. [15] Proposed
GMM-HMM GMM-HMM
(Persian ESD and CREMA-D)
DNN-HMM
(CREMA-D)
Features: MFCC
Emotions: Anger, Happy
Features: MFCC, M-MFCC, GFCC, and PNCC
Emotions: Anger, Disgust, Fear, Happy, and Sad
Features: ExpoLog, GFCC, and PNCC
Emotions: Anger, Disgust, Fear, Happy, and Sad
Warping > No Warping Warping > No Warping Warping > No Warping
DCT > Filterbank DCT > Filterbank DCT Filterbank
Filterbank & DCT > DCT Filterbank & DCT DCT Filterbank & DCT > DCT
Filterbank & DCT > Filterbank Filterbank & DCT Filterbank Filterbank & DCT Filterbank