| self-supervised-image-classification-on-imagenet-finetuned | dBOT (ViT-H/14) | #6 | Top 1 Accuracy: 88.0%Number of Params: 632M |
| instance-segmentation-on-coco-test-dev | dBOT ViT-L (CLIP) | #26 | mask AP: 48.8 |
| instance-segmentation-on-coco-test-dev | dBOT ViT-L | #29 | mask AP: 48.3 |
| instance-segmentation-on-coco-test-dev | dBOT ViT-B | #36 | mask AP: 46.3 |
| instance-segmentation-on-coco-test-dev | dBOT ViT-B (CLIP) | #37 | mask AP: 46.2 |
| object-detection-on-coco-test-dev | dBOT ViT-L (CLIP) | #36 | box mAP: 56.8 |
| object-detection-on-coco-test-dev | dBOT ViT-L | #41 | box mAP: 56.1 |
| object-detection-on-coco-test-dev | dBOT ViT-B | #60 | box mAP: 53.5 |
| semantic-segmentation-on-ade20k | dBOT ViT-L (CLIP) | #43 | Validation mIoU: 56.2 |
| semantic-segmentation-on-ade20k | dBOT ViT-L | #50 | Validation mIoU: 55.2 |
| semantic-segmentation-on-ade20k | dBOT ViT-B (CLIP) | #89 | Validation mIoU: 52.9 |
| semantic-segmentation-on-ade20k | dBOT ViT-B | #119 | Validation mIoU: 50.8 |
| image-classification-on-imagenet | dBOT ViT-H (CLIP as Teacher) | #62 | Top 1 Accuracy: 88.2% |
| image-classification-on-imagenet | dBOT ViT-L (CLIP as Teacher) | #77 | Top 1 Accuracy: 87.8% |
| image-classification-on-imagenet | dBOT ViT-B (CLIP as Teacher) | #221 | Top 1 Accuracy: 85.7% |
| object-detection-on-coco-val2017 | dBOT (ViT-B, CLIP teacher, Mask R-CNN) | #140 | box AP: 53.6 |