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

CAT

Benchmarks

Task NameDataset NameSOTA ResultTrend
Secret information extractionCat
Accuracy99.68
29
Image GenerationCAT 64x64 (N=9304)
Mean FID15.61
15
Surgical Video RecognitionCat-21
Top-1 Accuracy15.69
14
Image GenerationCAT 32x32 (train)
FID9.5
12
Video fittingCat
PSNR (dB)37.54
10
3D Reconstruction and RenderingCat
PSNR24.83
9
Interactive SegmentationCat
mIoU88.7
8
Surgical Video RecognitionCAT-1k
Top-1 Accuracy18.81
7
Surgical Video RecognitionCat-101
Top-1 Accuracy35.16
7
Image StitchingCat (test)
Inference Time (s)0.0389
7
Keyphrase RecommendationCAT 3
Precision54
5
Keyphrase RecommendationCAT 2
Precision56
5
Keyphrase RecommendationCAT 1
Precision54
5
3-D reconstructionCat Reflective Object (test)
RMSE (mm)0.881
5
Anti-steganalysisCat
YeNet Score0.9902
5
Blood Vessel ReconstructionCAT08 (1496)
Radius Difference0.101
4
Texture ReconstructionCat (test)
PSNR34.91
4
Image GenerationCAT 256x256 2011
Min FID32.11
4
Image GenerationCat
FID17.3
3
Image GenerationCAT 128x128
Min FID15.85
3
Gaussian DeblurringCat dataset
FID63.17
2
Controlled Image GenerationCat (test)
FID71.56
2
Text-to-Image GenerationCat
FID69.57
2
Showing 23 of 23 rows