Skip to main content

Table 2 The recall results of different combinations of wavelet-based features

From: Lightweight feature extraction method for efficient acoustic-based animal recognition in wireless acoustic sensor networks

Set Wavelet features selected Average recall using ED (%) Average recall using MD (%) Number of features Wavelet decomposition level(s)
#1 All features 88.82 89.86 37 5
#2 Entropy of level 2
Energy of level 1
Energy Ratio of level 2 & 1
Centroid of level 1
85.70 87.09 4 2
#3 Entropy of level 2
Energy of level 1
Energy Ratio of level 2 & 1
85.38 84.86 3 2
#4 STD of level 1
Energy of level 1
Entropy of level 1
86.04 86.87 3 1
#5 Energy of level 1
Entropy of level 1
86.08 85.59 2 1
#6 STD of level 1
Energy of level 1
85.03 85.44 2 1
#7 STD of level 1
Entropy of level 1
85.52 85.21 2 1