Swin Transformer (ICCV'2021)
@article{liu2021Swin,
title={Swin Transformer: Hierarchical Vision Transformer using Shifted Windows},
author={Liu, Ze and Lin, Yutong and Cao, Yue and Hu, Han and Wei, Yixuan and Zhang, Zheng and Lin, Stephen and Guo, Baining},
journal={arXiv preprint arXiv:2103.14030},
year={2021}
}
Segmentor | Pretrain | Backbone | Crop Size | Schedule | Train/Eval Set | mIoU | Download |
---|---|---|---|---|---|---|---|
UperNet | ImageNet-1k-224x224 | Swin-T | 512x512 | LR/POLICY/BS/EPOCH: 0.00006/poly/16/130 | train/val | 44.58% | cfg | model | log |
UperNet | ImageNet-1k-224x224 | Swin-S | 512x512 | LR/POLICY/BS/EPOCH: 0.00006/poly/16/130 | train/val | 48.39% | cfg | model | log |
UperNet | ImageNet-22k-384x384 | Swin-B | 512x512 | LR/POLICY/BS/EPOCH: 0.00006/poly/16/130 | train/val | 51.02% | cfg | model | log |
You can also download the model weights from following sources:
- BaiduNetdisk: https://pan.baidu.com/s/1gD-NJJWOtaHCtB0qHE79rA with access code s757