Skip to main content
. 2023 Jul 12;13:11257. doi: 10.1038/s41598-023-38190-x

Table 7.

Evaluation of the mask R-CNN model in high and low-density validation datasets, average mound features per image, after the entire DA workflow for the detection of hachure mound features.

Algorithm Density (%) TPs FNs FPs Recall (%) Precision (%) F1 (%)
High-density 128.26 77 38 3 66.96 96.25 78.97
Low-density 2.67 77 38 10 66.96 88.51 76.24