ImageNet
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
TinyImageNet
36.2Clean Accuracy
7
Tiny-ImageNet
53.65Accuracy (Original)
7
Tiny-ImageNet
8.96FGSM Error (eps=8/255)
7
ImageNet
17.688BD-Acc
7
ImageNet Ground Truth Class (val)
0.352Deletion Score
7
ImageNet Predicted Class (val)
37.79Deletion Score
7
ImageNet-1k (test)
62mAP (16 bits)
7
ImageNet-100 (test)
92.5mAP (16 bits)
7
ImageNet (val)
0.315L1 Loss
7
ImageNet 256 (test)
25.9PSNR
7
ImageNet-256 (test)
27.74PSNR
7
miniImageNet-C
75.37Session 0 Accuracy
7
ImageNet-A Inc-10 B-0 (test)
12.36Favg
7
ImageNet-A Inc-5 B-0 (test)
24.83Average Score
7
ImageNet 100 images (val)
82.5SSIM
7
TinyImageNet
25.51FPR95
7
TinyImageNet
41.4ID Accuracy
7
ImageNet-C Batch size=1
67.7Accuracy
7
ImageNet-A 10 steps incremental
63.19Average Accuracy
7
ImageNet
20.8PSNR
7
ImageNet
0.58LPIPS
7
ImageNet
0.58SSIM
7
ImageNet-P
89.96Accuracy (Gaussian Noise)
7
ImageNet (1000 images)
7Inference Time (s)
7
ImageNet (Source) to 9 Target Datasets (Caltech101, Pets, Cars, Flowers102, Aircraft, SUN397, DTD, Food101, UCF101) (test)
89.74Accuracy (Caltech101)
7