Skip to main content

Table 3 Performance achieved by the perceptual features with soft majority voting in comparison to majority voting

From: Perceptual audio features for emotion detection

   All (%) Arousal (%) Valence (%)
   UA WA UA WA UA WA
EMO-DB
   P-SVM S-MV 86.3 85.9 95.2 95.1 94.3 95.6
  MV 85.2 86.1 91.3 92.7 88.0 88.1
VAM
   P-SVM S-MV 61.3 76.2 69.4 71.7 59.9 83.3
  MV 57.1 74.7 59.2 73.3 52.4 77.8