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

CIFAR100

Benchmarks

Task NameDataset NameSOTA ResultTrend
Image ClassificationCIFAR100 (test)
Top-1 Accuracy93.4
407
Image ClassificationCIFAR100 long-tailed (test)
Accuracy66.3
155
Class Incremental LearningCIFAR100 (test)
Avg Acc78.45
116
Image classificationCIFAR100 (test)
Accuracy56.19
112
ClusteringCIFAR100 20
ACC61.4
93
Continual LearningCIFAR100 Split
Average Per-Task Accuracy95.38
85
OOD DetectionCIFAR100 Dfar
AUROC86.98
69
Image ClassificationCIFAR100 IDN (test)
Accuracy79.69
67
Class Incremental LearningCIFAR100 B50 (test)
Average Accuracy79.91
67
Continual LearningCIFAR100 Split 32x32 (test)
Accuracy30.1
66
Image ClassificationCIFAR100 UnLabel-Domain (UL)
Accuracy61.8
52
Image ClassificationCIFAR100 Labeled-Domain (L)
Accuracy76.4
52
Image ClassificationCIFAR100 10 tasks (test)
Accuracy49.7
51
Continual LearningCIFAR100 10 tasks (test)
Average Forgetting Rate5.6
51
Image ClassificationCIFAR100 (test)
Error Rate19.01
49
Image ClassificationCifar100 LT
F1 Score86.4
48
Class-Incremental LearningCIFAR100-LT rho=100 (test)
Avg Acc40.26
48
Semi-supervised Image ClassificationCIFAR100-LT (test)
Accuracy0.613
48
Partial Label LearningCIFAR100-LT
Accuracy64.75
48
Out-of-Distribution DetectionCIFAR100 ID Dnear OOD
AUROC81.37
47
Image ClassificationCIFAR100-LT (test)
Top-1 Acc (IR=100)59.59
45
Image ClassificationCIFAR100 without Cutout (test)
Accuracy79.71
45
OOD DetectionCIFAR100-C all 15 perturbations (test)
AUROC69.69
43
Image ClassificationCIFAR100 (test)
Accuracy92.21
43
Binary ClassificationBinary CIFAR100 (imbalance ratio 1:100) (test)
FPR @ 98% TPR64
41
Showing 25 of 206 rows
...