Method | Part A | Part B |
---|
MAE | RMSE | MAE | RMSE |
---|
MCNN [5] | 110.2 | 173.2 | 26.4 | 41.3 |
CP-CNN [25] | 73.6 | 106.4 | 20.1 | 30.1 |
ACSCP [24] | 75.7 | 102.7 | 17.2 | 27.4 |
CSRNet\(^\dagger\) [23] | 68.2 | 115.0 | 10.6 | 16.0 |
TasselNet [9] | 87.0 | 138.9 | 16.7 | 28.1 |
TasselNetv2 | 84.1 | 140.1 | 15.3 | 27.8 |
TasselNetv2\(^\dagger\) | 66.8 | 112.1 | 9.6 | 17.5 |
- \(^\dagger\) means the model is fine-tuned from the pretrained VGG16. The best performance is in italics