Table 2 Results of the deep learning model VGG16 on Urticaceae pollen from an area representing 10% of the total deposition area of Hirst-type aerobiological samples from Leiden (the Netherlands), Lleida and Vielha (both Catalonia, Spain).
From: Neural networks for increased accuracy of allergenic pollen monitoring
Sample location | Date collected | No. Pollen | % Urtica | % Parietaria | % U. mem | % Unknown | Identity threshold |
|---|---|---|---|---|---|---|---|
Leiden, NL | 23/08/2019 | 112 | 85.7 | 4.5 | 0 | 9.8 | 60% |
Lleida, SP | 16/06/2019 | 63 | 14.3 | 81.0 | 0 | 4.8 | 60% |
Vielha, SP | 09/08/2019 | 26 | 69.2 | 19.2 | 0 | 11.5 | 60% |
Leiden, NL | 23/08/2019 | 112 | 83.0 | 3.6 | 0 | 13.4 | 70% |
Lleida, SP | 16/06/2019 | 63 | 12.7 | 79.4 | 0 | 7.9 | 70% |
Vielha, SP | 09/08/2019 | 26 | 69.2 | 11.5 | 0 | 19.2 | 70% |
