Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

PanNuke

Benchmarks

Task NameDataset NameSOTA ResultTrend
Instance SegmentationPanNuke 19 tissue types (three-fold cross-validation)
mPQ59.9
120
Nuclei DetectionPanNuke averaged across three dataset splits
Precision0.88
40
Nuclei Instance SegmentationPanNuke
Neoplastic Score84.22
39
Nuclei SegmentationPanNuke T2
IoU69.3
28
Nuclei SegmentationPanNuke T1
mIoU70.36
28
Nuclei ClassificationPanNuke (official three-fold splits)
Precision (Neo)75
18
Nuclei ClassificationPanNuke
Neoplastic Precision74
15
Nuclei Instance SegmentationPanNuke (target)
Dice82.97
14
Panoptic SegmentationPanNuke (three-fold cross-validation)
Neoplastic58.4
12
classificationPanNuke
AUC91.5
11
Nuclei SegmentationPanNuke
Dice (All Nuclei)79.42
11
Medical Image SegmentationPanNuke
mIoU85.14
10
Histopathology Image ClassificationPanNuke
Accuracy71.52
10
Nuclei SegmentationPanNuKe
mPQ58.61
10
Binary Nuclei DetectionPanNuke
Precision84
10
Instance SegmentationPanNuke
MACs (G)89.5
8
Nuclei Panoptic SegmentationPanNuke (three-fold cross-val)
mPQ50.8
8
Nuclei instance segmentationPanNuke Average over all tissues
bPQ69.55
7
Nuclei instance segmentationPanNuke Uterus
Boundary PQ67.54
7
Nuclei instance segmentationPanNuke Thyroid
bPQ72.93
7
Nuclei instance segmentationPanNuke Testis
bPQ71.57
7
Nuclei instance segmentationPanNuke Stomach
bPQ0.7155
7
Nuclei instance segmentationPanNuke Skin
bPQ67.77
7
Nuclei instance segmentationPanNuke Prostate
bPQ69.85
7
Nuclei instance segmentationPanNuke Pancreatic
bPQ69.25
7
Showing 25 of 64 rows