-
-
Notifications
You must be signed in to change notification settings - Fork 108
/
Copy pathREADME.md
56 lines (39 loc) · 10.4 KB
/
README.md
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
## Introduction
<a href="https://github.com/zhanghang1989/ResNeSt">Official Repo</a>
<a href="https://github.com/SegmentationBLWX/sssegmentation/blob/main/ssseg/modules/models/backbones/resnest.py">Code Snippet</a>
<details>
<summary align="left"><a href="https://arxiv.org/pdf/2004.08955.pdf">ResNeSt (ArXiv'2020)</a></summary>
```latex
@article{zhang2020resnest,
title={ResNeSt: Split-Attention Networks},
author={Zhang, Hang and Wu, Chongruo and Zhang, Zhongyue and Zhu, Yi and Zhang, Zhi and Lin, Haibin and Sun, Yue and He, Tong and Muller, Jonas and Manmatha, R. and Li, Mu and Smola, Alexander},
journal={arXiv preprint arXiv:2004.08955},
year={2020}
}
```
</details>
## Results
#### PASCAL VOC
| Segmentor | Pretrain | Backbone | Crop Size | Schedule | Train/Eval Set | mIoU | Download |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
| FCN | ImageNet-1k-224x224 | S-101-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/60 | trainaug/val | 77.41% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/fcn_resnest101os8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/fcn_resnest101os8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/fcn_resnest101os8_voc.log) |
| PSPNet | ImageNet-1k-224x224 | S-101-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/60 | trainaug/val | 79.07% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/pspnet_resnest101os8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/pspnet_resnest101os8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/pspnet_resnest101os8_voc.log) |
| DeepLabV3 | ImageNet-1k-224x224 | S-101-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/60 | trainaug/val | 78.97% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/deeplabv3_resnest101os8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3_resnest101os8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3_resnest101os8_voc.log) |
| DeepLabV3plus | ImageNet-1k-224x224 | S-101-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/60 | trainaug/val | 79.76% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/deeplabv3plus_resnest101os8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3plus_resnest101os8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3plus_resnest101os8_voc.log) |
#### ADE20k
| Segmentor | Pretrain | Backbone | Crop Size | Schedule | Train/Eval Set | mIoU | Download |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
| FCN | ImageNet-1k-224x224 | S-101-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/130 | train/val | 45.74% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/fcn_resnest101os8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/fcn_resnest101os8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/fcn_resnest101os8_ade20k.log) |
| PSPNet | ImageNet-1k-224x224 | S-101-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/130 | train/val | 46.03% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/pspnet_resnest101os8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/pspnet_resnest101os8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/pspnet_resnest101os8_ade20k.log) |
| DeepLabV3 | ImageNet-1k-224x224 | S-101-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/130 | train/val | 46.24% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/deeplabv3_resnest101os8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3_resnest101os8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3_resnest101os8_ade20k.log) |
| DeepLabV3plus | ImageNet-1k-224x224 | S-101-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/130 | train/val | 46.48% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/deeplabv3plus_resnest101os8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3plus_resnest101os8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3plus_resnest101os8_ade20k.log) |
#### CityScapes
| Segmentor | Pretrain | Backbone | Crop Size | Schedule | Train/Eval Set | mIoU | Download |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
| FCN | ImageNet-1k-224x224 | S-101-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 78.14% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/fcn_resnest101os8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/fcn_resnest101os8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/fcn_resnest101os8_cityscapes.log) |
| PSPNet | ImageNet-1k-224x224 | S-101-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 78.70% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/pspnet_resnest101os8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/pspnet_resnest101os8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/pspnet_resnest101os8_cityscapes.log) |
| DeepLabV3 | ImageNet-1k-224x224 | S-101-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 79.75% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/deeplabv3_resnest101os8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3_resnest101os8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3_resnest101os8_cityscapes.log) |
| DeepLabV3plus | ImageNet-1k-224x224 | S-101-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 80.30% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/resnest/deeplabv3plus_resnest101os8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3plus_resnest101os8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_resnest/deeplabv3plus_resnest101os8_cityscapes.log) |
## More
You can also download the model weights from following sources:
- BaiduNetdisk: https://pan.baidu.com/s/1gD-NJJWOtaHCtB0qHE79rA with access code **s757**