CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows

Open paper
Benchmark Model Rank Results
semantic-segmentation-on-ade20k-valCSWin-L (UperNet, ImageNet-22k pretrain)#26mIoU: 55.7
semantic-segmentation-on-ade20kCSWin-L (UperNet, ImageNet-22k pretrain)#46Validation mIoU: 55.70
image-classification-on-imagenetCSWin-L (384 res,ImageNet-22k pretrain)#87Top 1 Accuracy: 87.5%Number of params: 173MGFLOPs: 96.8

Rank counts only results with a code link.