Table 3 Comparison of different losses on RFRB
From: Automatic rape flower cluster counting method based on low-cost labelling and UAV-RGB images
Backbone | Loss | Acc | MAE | rMAE | rMSE | rrMSE | \(R^2\) |
|---|---|---|---|---|---|---|---|
Vgg19 | OT | 0.6499 | 108.07 | 10.40 | 128.08 | 38.77 | 0.6492 |
TV | 0.7132 | 82.91 | 9.12 | 97.61 | 30.16 | 0.2587 | |
MAE | 0.8981 | 25.94 | 5.09 | 34.43 | 12.56 | 0.9512 | |
MSE | 0.9027 | 24.40 | 4.94 | 32.91 | 11.70 | 0.9534 | |
DM | 0.9014 | 24.43 | 4.94 | 32.38 | 12.99 | 0.9589 | |
Bayesian | 0.9098 | 22.90 | 4.78 | 31.95 | 12.15 | 0.9623 | |
RapeNet+ | OT | 0.8306 | 45.31 | 6.73 | 60.69 | 19.75 | 0.8841 |
TV | 0.4787 | 149.05 | 12.21 | 168.70 | 52.71 | – | |
MAE | 0.8733 | 37.06 | 6.09 | 49.67 | 14.32 | 0.8658 | |
MSE | 0.8896 | 29.96 | 5.47 | 42.27 | 13.47 | 0.9428 | |
DM | 0.9017 | 23.10 | 4.81 | 32.08 | 12.28 | 0.9613 | |
Bayesian | 0.9026 | 23.65 | 4.86 | 29.95 | 12.03 | 0.9635 |