Skip to main content

Table 3 Original features

From: Emotion recognition and school violence detection from children speech

Original feature contour

f1

RMSenergy root-mean-square signal frame energy

f2–f13

Mel-frequency cepstrum coefficients 1–12

f14

Zero-crossing rate of time signal (frame-based)

f15

The voicing probability computed from the ACF

f16

The fundamental frequency computed from the cepstrum