| Benchmark | Model | Rank | Results |
|---|---|---|---|
| semantic-segmentation-on-densepass | PoolFormer (MiT-B1) | #9 | mIoU: 43.18% |
| instance-segmentation-on-coco-minival | PoolFormer-S36 + Mask R-CNN | #103 | mask AP: 37.7 |
| semantic-segmentation-on-ade20k | PoolFormer-M48 | #228 | Validation mIoU: 42.7 |
| object-detection-on-coco-val2017 | PoolFormer-S36 (Mask R-CNN) | #380 | box AP: 41.0AP50: 63.1AP75: 44.8 |
| image-classification-on-imagenet | MetaFormer PoolFormer-M48 | #565 | Top 1 Accuracy: 82.5%Number of params: 73MGFLOPs: 23.2 |