| object-detection-on-coco-test-dev | FocalNet-H (DINO) | #8 | box mAP: 64.4 |
| object-detection-on-coco-val2017 | FocalNet-H (DINO) | #10 | box AP: 64.2 |
| object-detection-on-coco-val2017 | FocalNet-T (LRF, Cascade Mask R-CNN) | #186 | box AP: 51.5AP50: 70.3AP75: 56.0 |
| object-detection-on-coco-val2017 | FocalNet-T (SRF, Cascade Mask R-CNN) | #447 | AP50: 70.1AP75: 55.8 |
| panoptic-segmentation-on-coco-minival | FocalNet-L (Mask2Former (200 queries)) | #16 | PQ: 57.9AP: 48.4 |
| semantic-segmentation-on-ade20k | FocalNet-L (Mask2Former) | #21 | Validation mIoU: 58.5 |
| instance-segmentation-on-coco-minival | FocalNet-B + Mask R-CNN | #72 | mask AP: 44.1 |
| image-classification-on-imagenet | FocalNet-L | #105 | Top 1 Accuracy: 87.3% |