Anomaly Detection on Anomaly Detection on Unlabeled ImageNet-30 vs CUB-200

Metric: ROC-AUC (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Hide augmentations

Sort:

#	Model↕	ROC-AUC▼	Augmentations	Paper	Date↕	Code
1	PsudoLabels CLIP ViT	99.4	Yes	Out-of-Distribution Detection Without Class Labels	2021-12-14	-
2	DN2 CLIP ViT	93.8	Yes	Deep Nearest Neighbor Anomaly Detection	2020-02-24	-
3	GOAD	90.5	No	Classification-Based Anomaly Detection for Gener...	2020-05-05	Code
4	ROT+Trans	74.5	No	Using Self-Supervised Learning Can Improve Model...	2019-06-28	Code
5	CSI	71.5	No	CSI: Novelty Detection via Contrastive Learning ...	2020-07-16	Code

#1PsudoLabels CLIP ViTSOTA
99.4
ROC-AUC· Augmentations· 2021-12-14
Out-of-Distribution Detection Without Class Labels
#2DN2 CLIP ViTSOTA
93.8
ROC-AUC· Augmentations· 2020-02-24
Deep Nearest Neighbor Anomaly Detection
#3GOAD
90.5
ROC-AUC· 2020-05-05
Classification-Based Anomaly Detection for General Data Code
#4ROT+TransSOTA
74.5
ROC-AUC· 2019-06-28
Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty Code
#5CSI
71.5
ROC-AUC· 2020-07-16
CSI: Novelty Detection via Contrastive Learning on Distributionally Shifted Instances Code