Table 4 Common acoustic features obtained by applying ANOVA, RF and mRMR feature selection methods on the recorded audio files of PDT and SRT
PD Task | SR task |
|---|---|
MFCC 13 (mean) | MFCC 12,13 (skew) |
MFCC 12,13 (kurt) | ΔMFCC 3,4,6,13 (mean) |
MFCC 10,13 (skew) | ΔMFCC 4 (skew) |
ΔMFCC 2,11 (mean) | ΔLSP freq 7 (mean) |
ΔMFCC 2,3,6,7 (kurt) | ΔLSP freq 3,6 (kurt) |
ΔMFCC 6,11,13 (skew) | Loudness (kurt, skew) |
ΔLSP freq 3,5 (mean) | ΔLoudness (kurt, skew) |
ΔLSP freq 2,6 (kurt) | F0 (kurt) |
ΔLSP freq 1 (skew) | ΔF0 (mean) |