Weakly Supervised Instance Segmentation On 2
評価指標
AP
AP@50
AP@75
AP@L
AP@M
AP@S
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
比較表
| モデル名 | AP | AP@50 | AP@75 | AP@L | AP@M | AP@S |
|---|---|---|---|---|---|---|
| boxinst-high-performance-instance | 32.1 | 55.1 | 32.4 | 43.5 | 34.3 | 15.6 |
| discobox-weakly-supervised-instance | 35.8 | 59.8 | 36.4 | 52.1 | 38.7 | 16.9 |
| boxinst-high-performance-instance | 33.9 | 57.7 | 34.5 | 46.6 | 36.1 | 16.5 |
| discobox-weakly-supervised-instance | 37.9 | 61.4 | 40.0 | 53.9 | 41.1 | 18.0 |
| boxinst-high-performance-instance | 33.2 | 56.5 | 33.6 | 45.1 | 35.3 | 16.2 |
| boxinst-high-performance-instance | 35.0 | 59.3 | 35.6 | 48.9 | 37.2 | 17.1 |
| discobox-weakly-supervised-instance | 32.0 | 53.6 | 32.6 | 48.4 | 33.7 | 11.7 |