| Paper | Code | Average Best Overlap | mAP@0.25 | mAP@0.5 | mAP@0.75 | ModelName | ReleaseDate |
|---|---|---|---|---|---|---|---|
| BBAM: Bounding Box Attribution Map for Weakly Supervised Semantic and Instance Segmentation | ✓ Link | 63.0 | 76.8 | 63.7 | 31.8 | BBAM | 2021-03-16 |
| Leveraging Instance-, Image- and Dataset-Level Information for Weakly Supervised Instance Segmentation | ✓ Link | 50.8 | - | 48.4 | 24.9 | LIID | 2020-09-10 |
| Weakly Supervised Instance Segmentation by Deep Community Learning | 48.2 | 56.6 | 38.1 | 12.3 | WSIS_CL | 2020-01-30 | |
| Complete Instances Mining for Weakly Supervised Instance Segmentation | ✓ Link | 68.7 | 55.9 | 30.9 | CIM + Mask R-CNN | 2024-02-12 | |
| Beyond Semantic to Instance Segmentation: Weakly-Supervised Instance Segmentation via Semantic Knowledge Transfer and Self-Refinement | ✓ Link | - | 66.4 | 56.1 | 30.2 | BESTIE (point label, proposal-free) | 2021-09-20 |
| Beyond Semantic to Instance Segmentation: Weakly-Supervised Instance Segmentation via Semantic Knowledge Transfer and Self-Refinement | ✓ Link | - | 61.2 | 51.0 | 26.6 | BESTIE (image-level label, proposal-free) | 2021-09-20 |