Table 4 Performance of different models
From: Towards accurate bird sound recognition through multi-scale texture-aware modeling
Model | AR (%) | PR (%) | RR (%) | F1-S (%) |
|---|---|---|---|---|
DLoGNet | 91.18 | 91.09 | 91.23 | 91.16 |
CNN | 87.82 | 87.75 | 87.82 | 87.79 |
LSTM | 87.64 | 87.55 | 87.64 | 87.60 |
CNN-LSTM | 90.41 | 90.30 | 90.36 | 90.33 |
EfficientNet | 89.82 | 89.69 | 89.80 | 89.75 |
VGG-16 | 90.52 | 90.43 | 90.50 | 90.47 |
Transformer | 91.18 | 91.05 | 91.26 | 91.16 |
MDF-Net | 91.16 | 91.09 | 91.13 | 91.11 |
