Development of the Arabic Voice Pathology Database and Its Evaluation by Using Speech Features and Machine Learning Algorithms
Table 8
Overall best accuracies (%) for sustained vowels and running speech by using the MEEI database.
Features
Experiments
SVM
GMM
VQ
HMM
/AH/
Rainbow
/AH/
Rainbow
/AH/
Rainbow
/AH/
Rainbow
MFCC
Detection
97.4
91.6
97.3
90.3
96.0
88.9
Classification
95.4
97.3
96.3
97.3
87.5
88.9
LPCC
Detection
97.9
90.7
96.4
83.2
97.8
87.6
Classification
95.4
97.3
97.3
97.3
87.5
97.3
RASTA-PLP
Detection
98.0
91.6
84.1
96.4
88.9
98.1
Classification
95.5
97.3
97.3
96.3
85.2
84.6
LPC
Detection
82.9
96.0
78.3
97.3
80.1
96.3
Classification
95.2
97.3
97.3
94.4
75.0
82.5
PLP
Detection
87.8
96.8
89.4
87.4
96.3
Classification
95.0
97.3
97.3
94.4
61.1
84.6
MDVP
Detection
—
88.3
—
68.3
—
—
—
Classification
—
—
—
—
—
—
—
best detection rate for sustained vowels. The best detection rate for running speech. The best classification rate for sustained vowels. The best classification rate for running speech.
