Anomaly Detection on Leave-One-Class-Out ImageNet-30

Metric: AUROC (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Hide augmentations

Sort:

#	Model↕	AUROC▼	Augmentations	Paper	Date↕	Code
1	BCE-CLIP (OE)	99.3	Yes	Exposing Outlier Exposure: What Can Be Learned F...	2022-05-23	Code
2	CLIP (zero shot)	97.8	No	Exposing Outlier Exposure: What Can Be Learned F...	2022-05-23	Code
3	DSAD	88.8	Yes	Exposing Outlier Exposure: What Can Be Learned F...	2022-05-23	Code
4	HSC (OE)	88.3	Yes	Exposing Outlier Exposure: What Can Be Learned F...	2022-05-23	Code
5	Binary Cross Entropy (OE)	88.2	Yes	Exposing Outlier Exposure: What Can Be Learned F...	2022-05-23	Code
6	DSVDD	49.7	No	Exposing Outlier Exposure: What Can Be Learned F...	2022-05-23	Code

#1BCE-CLIP (OE)SOTA
99.3
AUROC· Augmentations· 2022-05-23
Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images Code
#2CLIP (zero shot)
97.8
AUROC· 2022-05-23
Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images Code
#3DSAD
88.8
AUROC· Augmentations· 2022-05-23
Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images Code
#4HSC (OE)
88.3
AUROC· Augmentations· 2022-05-23
Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images Code
#5Binary Cross Entropy (OE)
88.2
AUROC· Augmentations· 2022-05-23
Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images Code
#6DSVDD
49.7
AUROC· 2022-05-23
Exposing Outlier Exposure: What Can Be Learned From Few, One, and Zero Outlier Images Code