Good feature for speech based Emotion Identification

DYNAMIC:         Short frame, sentence level feature

STATIC:

f0
duration
MFB:  Mel filter Bank coefficent
Voice Quality


Other:
JItter, HNR, shimmer