-
-
Notifications
You must be signed in to change notification settings - Fork 108
/
Copy pathREADME.md
37 lines (24 loc) · 3.55 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
## Introduction
<a href="https://github.com/wuhuikai/FastFCN">Official Repo</a>
<a href="https://github.com/SegmentationBLWX/sssegmentation/blob/main/ssseg/modules/models/segmentors/fastfcn/fastfcn.py">Code Snippet</a>
<details>
<summary align="left"><a href="https://arxiv.org/pdf/1903.11816.pdf">FastFCN (ArXiv'2019)</a></summary>
```latex
@article{wu2019fastfcn,
title={Fastfcn: Rethinking dilated convolution in the backbone for semantic segmentation},
author={Wu, Huikai and Zhang, Junge and Huang, Kaiqi and Liang, Kongming and Yu, Yizhou},
journal={arXiv preprint arXiv:1903.11816},
year={2019}
}
```
</details>
## Results
#### CityScapes
| Segmentor | Pretrain | Backbone | Crop Size | Schedule | Train/Eval Set | mIoU | Download |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
| EncNet | ImageNet-1k-224x224 | R-50-D32 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 78.42% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/fastfcn/fastfcn_encnet_resnet50os32_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_fastfcn/fastfcn_encnet_resnet50os32_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_fastfcn/fastfcn_encnet_resnet50os32_cityscapes.log) |
| PSPNet | ImageNet-1k-224x224 | R-50-D32 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 79.36% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/fastfcn/fastfcn_pspnet_resnet50os32_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_fastfcn/fastfcn_pspnet_resnet50os32_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_fastfcn/fastfcn_pspnet_resnet50os32_cityscapes.log) |
| DeepLabV3 | ImageNet-1k-224x224 | R-50-D32 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 79.96% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/fastfcn/fastfcn_deeplabv3_resnet50os32_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_fastfcn/fastfcn_deeplabv3_resnet50os32_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_fastfcn/fastfcn_deeplabv3_resnet50os32_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**