Table 2 Performance comparison of Model-2, Model-K, and Count-ception
From: Counting sea lions and elephants from aerial photography using deep learning with density maps
Model | Feature extractor | SSL | Elephant | Parameters | ||
|---|---|---|---|---|---|---|
RMSE | MAE | RMSE | MAE | |||
Model-2 | Eff.Net-B5 | 1.88 | 1.09 | 0.60 | 0.34 | \(\approx\)37M |
Model-K | VGG | 2.17 | 1.43 | 0.81 | 0.43 | \(\approx\)48M |
Count-ception | No | 5.57 | 3.54 | 1.59 | 0.84 | \(\approx\)14M |