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

Kvasir

Benchmarks

Task NameDataset NameSOTA ResultTrend
Polyp SegmentationKvasir
Dice Score93.6
143
Medical Image SegmentationKvasir-Seg
Dice Score92.17
98
Polyp SegmentationKvasir (test)
Dice Coefficient97.4
82
Medical Image SegmentationKvasir-SEG
Dice Coefficient88.79
75
Image ClassificationKvasir
Mean Accuracy94.3
63
Medical Image SegmentationKvasir
mDice95.7
60
Semantic SegmentationKvasir-SEG (test)
IoU810.34
60
Polyp SegmentationKvasir-SEG
mIoU0.879
54
OOD DetectionKvasir v2
AUROC94.55
37
Medical Image ClassificationKvasir
Accuracy92.84
37
Anomaly LocalizationKvasir
AUROC95
22
Anomaly DetectionKvasir
Pixel AUROC90.6
22
Anomaly SegmentationKvasir
AP70.6
20
Polyp SegmentationKvasir Noisy (test)
Dice Coefficient85.3
18
Polyp SegmentationKvasir Clean (test)
Dice Coefficient89.6
18
Polyp SegmentationKvasir source: CVC target domain
DSC0.921
18
Colon Polyp SegmentationKvasir (5-fold cross-validation)
Dice Score92.6
18
Medical Image SegmentationKvasir (test)
F1 Score93.9
18
Surgical Video RecognitionKvasir-C
Top-1 Accuracy73.17
14
2D Medical Image SegmentationKvasir
ASD9.36
14
Polyp SegmentationKvasir 33 19 (test)
Dice Coefficient91.8
14
Polyp SegmentationKvasir (val)
Dice Coefficient90.3
14
Semantic SegmentationKvasir-SEG
mIoU89.03
13
Polyp segmentationKvasir In-Domain
Dice90.19
13
Medical Image SegmentationKvasir-Sessile (test)
mIoU84.6
12
Showing 25 of 55 rows