| image-classification-on-inaturalist-2018 | InternImage-H | #1 | Top-1 Accuracy: 92.6% |
| image-classification-on-places205 | InternImage-H | #1 | Top 1 Accuracy: 71.7% |
| image-classification-on-places365 | InternImage-H（CNN） | #1 | Top 1 Accuracy: 61.2% |
| object-detection-on-crowdhuman-full-body | InternImage-H | #1 | AP: 97.2 |
| object-detection-on-pascal-voc-2012 | InternImage-H | #1 | MAP: 97.2 |
| object-detection-on-coco-test-dev | InternImage-H | #2 | box mAP: 65.4Params (M): 2180 |
| object-detection-on-coco-test-dev | InternImage-XL | #9 | box mAP: 64.3Params (M): 602 |
| semantic-segmentation-on-pascal-context | InternImage-H | #2 | mIoU: 70.3 |
| semantic-segmentation-on-replica | InternImage | #2 | mIoU: 38.4 |
| instance-segmentation-on-coco-minival | InternImage-H | #3 | mask AP: 55.4AP50: 80.1AP75: 61.5APL: 74.4APM: 58.4APS: 37.9 |
| instance-segmentation-on-coco-minival | InternImage-XL | #35 | mask AP: 48.8GFLOPs: 1782Params (M): 387 |
| instance-segmentation-on-coco-minival | InternImage-L | #40 | mask AP: 48.5GFLOPs: 1399Params (M): 277box AP: 56.1 |
| instance-segmentation-on-coco-minival | InternImage-S | #67 | mask AP: 44.5GFLOPs: 340Params (M): 69box AP: 49.7 |
| instance-segmentation-on-coco-minival | InternImage-T | #75 | mask AP: 43.7GFLOPs: 270Params (M): 49box AP: 49.1 |
| instance-segmentation-on-coco-minival | InternImage-B | #114 | GFLOPs: 501Params (M): 115 |
| object-detection-on-lvis-v1-0-minival | InternImage-H | #3 | box AP: 65.8 |
| object-detection-on-lvis-v1-0-val | InternImage-H | #3 | box AP: 63.2 |
| semantic-segmentation-on-ade20k | InternImage-H | #3 | Validation mIoU: 62.9Params (M): 1310GFLOPs: 4635 |
| semantic-segmentation-on-ade20k | InternImage-XL | #49 | Validation mIoU: 55.3Params (M): 368GFLOPs: 3142 |
| semantic-segmentation-on-ade20k | InternImage-L | #69 | Validation mIoU: 54.1Params (M): 256GFLOPs: 2526 |
| semantic-segmentation-on-ade20k | InternImage-B | #108 | Validation mIoU: 51.3Params (M): 128GFLOPs: 1185 |
| semantic-segmentation-on-ade20k | InternImage-S | #118 | Validation mIoU: 50.9Params (M): 80GFLOPs: 1017 |
| semantic-segmentation-on-ade20k | InternImage-T | #169 | Validation mIoU: 48.1Params (M): 59GFLOPs: 944 |
| semantic-segmentation-on-ade20k | InternImage-H (M3I Pre-training) | #247 | Params (M): 1310 |
| semantic-segmentation-on-cityscapes-test | InternImage-H | #3 | Mean IoU (class): 86.1% |
| semantic-segmentation-on-cityscapes-val | InternImage-H | #4 | mIoU: 87 |
| semantic-segmentation-on-cityscapes-val | InternImage-XL | #6 | mIoU: 86.4 |
| object-detection-on-coco-val2017 | InternImage-H | #5 | box AP: 65.0 |
| object-detection-on-coco-val2017 | InternImage-XL | #11 | box AP: 64.2 |
| object-detection-on-coco-o | InternImage-L (Cascade Mask R-CNN) | #8 | Average mAP: 37.0Effective Robustness: 11.72 |
| image-classification-on-imagenet | InternImage-H | #17 | Top 1 Accuracy: 89.6%Number of params: 1080MGFLOPs: 1478 |
| image-classification-on-imagenet | InternImage-XL | #68 | Top 1 Accuracy: 88%Number of params: 335MGFLOPs: 163 |
| image-classification-on-imagenet | InternImage-L | #82 | Top 1 Accuracy: 87.7%Number of params: 223MGFLOPs: 108 |
| image-classification-on-imagenet | InternImage-B | #297 | Top 1 Accuracy: 84.9%Number of params: 97MGFLOPs: 16 |
| image-classification-on-imagenet | InternImage-S | #366 | Top 1 Accuracy: 84.2%Number of params: 50MGFLOPs: 8 |
| image-classification-on-imagenet | InternImage-DCNv3-G (M3I Pre-training) | #1109 | Number of params: 3000M |
| instance-segmentation-on-coco-test-dev | InternImage-H | #104 | AP50: 80.8AP75: 62.2APS: 41.0APM: 58.9APL: 70.3 |