Classification on ImageNet C-OOD (class-out-of-distribution)

Metric: Detection AUROC (severity 5) (higher is better)

LeaderboardDataset

Loading chart...

Results

Submit a result

Sort:

#	Model↕	Detection AUROC (severity 5)▼	Augmentations	Paper	Date↕	Code
1	ViT-L/32-384 with Max-logit	0.9632	No	A framework for benchmarking class-out-of-distri...	2023-02-23	Code
2	ViT-L/32-384 with ODIN	0.9589	No	A framework for benchmarking class-out-of-distri...	2023-02-23	Code
3	ViT-L/32-384 with Entropy	0.9514	No	A framework for benchmarking class-out-of-distri...	2023-02-23	Code
4	ViT-L/32-384 with MC Dropout	0.9478	No	A framework for benchmarking class-out-of-distri...	2023-02-23	Code
5	ViT-L/32-384 with Softmax	0.9293	No	A framework for benchmarking class-out-of-distri...	2023-02-23	Code

#1ViT-L/32-384 with Max-logitSOTA
0.9632
Detection AUROC (severity 5)· 2023-02-23
A framework for benchmarking class-out-of-distribution detection and its application to ImageNet Code
#2ViT-L/32-384 with ODIN
0.9589
Detection AUROC (severity 5)· 2023-02-23
A framework for benchmarking class-out-of-distribution detection and its application to ImageNet Code
#3ViT-L/32-384 with Entropy
0.9514
Detection AUROC (severity 5)· 2023-02-23
A framework for benchmarking class-out-of-distribution detection and its application to ImageNet Code
#4ViT-L/32-384 with MC Dropout
0.9478
Detection AUROC (severity 5)· 2023-02-23
A framework for benchmarking class-out-of-distribution detection and its application to ImageNet Code
#5ViT-L/32-384 with Softmax
0.9293
Detection AUROC (severity 5)· 2023-02-23
A framework for benchmarking class-out-of-distribution detection and its application to ImageNet Code