Zero-shot Image Classification on ImageNet-A (test)
0.705AccuracyCLIP
Evaluation results
| Method | Links | |
|---|---|---|
| CLIPBackbone=ViT-L/14, Pretrain Data=WIT-400M, Evaluation Protocol=Zero-shot2023.01 | 0.705 | |
| REACT (Locked-Text Gated-Image)Backbone=ViT-L/14, Pretrain Data=WIT-400M, Evaluation Protocol=Zero-shot2023.01 | 0.686 | |
| REACT (Locked-Text Gated-Image)Backbone=ViT-L/14, Pretrain Data=LAION-2B, Evaluation Protocol=Zero-shot2023.01 | 0.552 | |
| CLIPBackbone=ViT-B/16, Pretrain Data=WIT-400M, Evaluation Protocol=Zero-shot2023.01 | 0.497 | |
| REACT (Locked-Text Gated-Image)Backbone=ViT-B/16, Pretrain Data=WIT-400M, Evaluation Protocol=Zero-shot2023.01 | 0.485 | |
| OpenCLIPBackbone=ViT-L/14, Pretrain Data=LAION-2B, Evaluation Protocol=Zero-shot2023.01 | 0.42 | |
| REACT (Locked-Text)Backbone=ViT-B/16, Pretrain Data=WIT-400M, Evaluation Protocol=Zero-shot2023.01 | 0.388 | |
| REACT (Locked-Text Gated-Image)Backbone=ViT-B/16, Pretrain Data=LAION-400M, Evaluation Protocol=Zero-shot2023.01 | 0.378 | |
| REACT (Locked-Text)Backbone=ViT-B/16, Pretrain Data=LAION-400M, Evaluation Protocol=Zero-shot2023.01 | 0.337 | |
| OpenCLIPBackbone=ViT-B/16, Pretrain Data=LAION-400M, Evaluation Protocol=Zero-shot2023.01 | 0.33 | |
| CLIPBackbone=ViT-B/32, Pretrain Data=WIT-400M, Evaluation Protocol=Zero-shot2023.01 | 0.314 | |
| REACT (Locked-Text Gated-Image)Backbone=ViT-B/32, Pretrain Data=WIT-400M, Evaluation Protocol=Zero-shot2023.01 | 0.308 | |
| REACT (Locked-Text Gated-Image)Backbone=ViT-B/32, Pretrain Data=LAION-2B, Evaluation Protocol=Zero-shot2023.01 | 0.279 | |
| OpenCLIPBackbone=ViT-B/32, Pretrain Data=LAION-2B, Evaluation Protocol=Zero-shot2023.01 | 0.262 | |
| REACT (Locked-Text)Backbone=ViT-B/32, Pretrain Data=LAION-2B, Evaluation Protocol=Zero-shot2023.01 | 0.238 | |
| REACT (Locked-Text)Backbone=ViT-B/32, Pretrain Data=WIT-400M, Evaluation Protocol=Zero-shot2023.01 | 0.23 | |
| REACT (Locked-Text Gated-Image)Backbone=ViT-B/32, Pretrain Data=LAION-400M, Evaluation Protocol=Zero-shot2023.01 | 0.227 | |
| OpenCLIPBackbone=ViT-B/32, Pretrain Data=LAION-400M, Evaluation Protocol=Zero-shot2023.01 | 0.218 | |
| REACT (Locked-Text)Backbone=ViT-B/32, Pretrain Data=LAION-400M, Evaluation Protocol=Zero-shot2023.01 | 0.202 |