FOR-instance
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
FOR-instance V2 (test)
91.4Precision
11
FOR-instance (SCION, New Zealand) V2
97.4Precision (P)
10
FOR-instance v2 (test)
78.2IoU @ 1 click
6
FOR-instance v2 (test)
99.1IoU (Ground)
6
FOR-instance Weighted Mean V2
91.4Precision
5
FOR-instance (TU-WIEN)
69.4F1 Score
4
FOR-instance (SCION site)
91.6F1 Score
4
FOR-instance RMIT
83F1 Score
4
FOR-instance (NIBIO site)
91.9F1 Score
4
FOR-instance (CULS site)
100F1 Score
4
FOR-instance Weighted Mean V2
96.5IoUG
4
FOR-instance NIBIO, Norway V2
96.2IoU (Ground)
3
FOR-instance NIBIO Norway V2
94.3Precision
3
FOR-instance BlueCat, Czech Republic V2
65.8IoU (W)
3
FOR-instance BlueCat, Czech Republic V2
87.6Precision (P)
3
FOR-instance RMIT, Australia V2
97.7IoUG
3
FOR-instance RMIT, Australia V2
89.3Precision
3
FOR-instance SCION, New Zealand V2
99.6IoUG
3
FOR-instance TUWIEN, Austria V2
98.8IoU (G)
3
FOR-instance TUWIEN, Austria V2
91.7Precision
3
FOR-instance YuChen, French Guiana V2
99.8IoUG
3
FOR-instance YuChen French Guiana V2
88.9Precision
3
FOR-instance CULS, Czech Republic V2
99.8IoU G
3
FOR-instance CULS Czech Republic V2
100Precision
3