Image Level Supervised Instance Segmentation

Results

Performance results of various models on this benchmark

Model Name	[email protected]	[email protected]	[email protected]	Paper Title	Repository
WISE	49.2	41.7	23.7	Where are the Masks: Instance Segmentation with Image-level Supervision
WeakSAM-Mask RCNN (with SAM)	70.3	59.6	36.2	WeakSAM: Segment Anything Meets Weakly-supervised Instance-level Recognition
IAM	45.9	28.3	11.9	Learning Instance Activation Maps for Weakly Supervised Instance Segmentation	-
Label-PEnet	49.2	30.2	12.9	Label-PEnet: Sequential Label Propagation and Enhancement Networks for Weakly Supervised Instance Segmentation	-
Arun et al.	59.7	50.9	28.5	Weakly Supervised Instance Segmentation by Learning Annotation Consistent Instances	-
IRN (proposal-free)	-	46.7	-	Weakly Supervised Learning of Instance Segmentation with Inter-pixel Relations
WeakSAM-Mask2Former (with SAM)	73.4	64.4	45.3	WeakSAM: Segment Anything Meets Weakly-supervised Instance-level Recognition
CIM + Mask R-CNN	68.7	55.9	30.9	Complete Instances Mining for Weakly Supervised Instance Segmentation
PRM	44.3	26.8	9.0	Weakly Supervised Instance Segmentation using Class Peak Response
CL	56.6	38.1	12.3	Weakly Supervised Instance Segmentation by Deep Community Learning	-
CIM	64.9	51.1	26.1	Complete Instances Mining for Weakly Supervised Instance Segmentation
LIID	-	48.4	24.9	Leveraging Instance-, Image- and Dataset-Level Information for Weakly Supervised Instance Segmentation
BESTIE (proposal-free)	61.2	51.0	26.6	Beyond Semantic to Instance Segmentation: Weakly-Supervised Instance Segmentation via Semantic Knowledge Transfer and Self-Refinement

0 of 13 row(s) selected.