Image Classification on ImageNet IN-1K IN-21K (val)
94.9AccuracyOpenCLIP-h
Evaluation results
| Method | Links | |
|---|---|---|
| OpenCLIP-hNum. Params=632M, Pretraining Dataset=LAION-2B, DG method=Text-guided pretrain2025.03 | 94.9 | |
| DINOv2-gNum. Params=1.1B, Pretraining Dataset=LVD-142M, DG method=Dataset curation2025.03 | 94.7 | |
| OpenCLIP-bNum. Params=87M, Pretraining Dataset=LAION-2B, DG method=Text-guided pretrain2025.03 | 94.3 | |
| DINOv2-bNum. Params=90M, Pretraining Dataset=LVD-142M, DG method=Dataset curation2025.03 | 93.6 | |
| SwinV2-BNum. Params=88M, Pretraining Dataset=IN-1K, DG method=-2025.03 | 91.9 | |
| SwinV2-SNum. Params=50M, Pretraining Dataset=IN-1K, DG method=-2025.03 | 91.7 | |
| SwinV2-TNum. Params=28M, Pretraining Dataset=IN-1K, DG method=-2025.03 | 90.6 | |
| EfficientNet-B0Num. Params=5M, Pretraining Dataset=IN-1K, DG method=-2025.03 | 88.2 | |
| EfficientNet-B3Num. Params=12M, Pretraining Dataset=IN-1K, DG method=-2025.03 | 88.1 | |
| ResNet-152Num. Params=60M, Pretraining Dataset=IN-1K, DG method=-2025.03 | 87.8 | |
| ResNet-50Num. Params=26M, Pretraining Dataset=IN-21K, DG method=DeepAugment*+AugMix†2025.03 | 87 | |
| ResNet-50Num. Params=26M, Pretraining Dataset=IN-1K, DG method=-2025.03 | 86 |