Update README.md

README.md

@@ -36,7 +36,7 @@ size_categories:
 ## Dataset Summary
 
 
-The **ImageNet-Hard** is a new benchmark that comprises 13,630 images, collected from various existing ImageNet-scale benchmarks (ImageNet, ImageNet-Sketch, ImageNet-C, ImageNet-R, ImageNet-ReaL, ImageNet-A, and ObjectNet). 
+The **ImageNet-Hard** is a new benchmark that comprises 11,350 images, collected from various existing ImageNet-scale benchmarks (ImageNet, ImageNet-V2, ImageNet-Sketch, ImageNet-C, ImageNet-R, ImageNet-ReaL, ImageNet-A, and ObjectNet). 
 This dataset is challenging to state-of-the-art vision models, as merely zooming in often fails to enhance their ability to classify images correctly. Consequently, even the most advanced models, such as `CLIP-ViT-L/14@336px`, struggle to perform well on this dataset, achieving only `2.02%` accuracy.
 
 
@@ -49,16 +49,16 @@ This dataset is challenging to state-of-the-art vision models, as merely zooming
 
 
 | Model               | Accuracy |
-| ------------------- | ----- |
-| ResNet-18           | 9.41  |
-| ResNet-50           | 12.56 |
-| ViT-B/32            | 15.95 |
-| VGG19               | 10.32 |
-| AlexNet             | 6.35  |
-| CLIP-ViT-L/14@224px | 1.86  |
-| CLIP-ViT-L/14@336px | 2.02  |
-| EfficientNet-L2-Ns  | 34.23 |
-
+| ------------------- | -------- |
+| ResNet-18           | 11.11    |
+| ResNet-50           | 14.91    |
+| ViT-B/32            | 18.78    |
+| VGG19               | 12.15    |
+| AlexNet             | 7.30     |
+| EfficientNet-B7     | 18.02    |
+| EfficientNet-L2-Ns  | 38.79    |
+| CLIP-ViT-L/14@224px | 2.11     |
+| CLIP-ViT-L/14@336px | 2.30     |
 
 **Evaluation Code**