-
-
Notifications
You must be signed in to change notification settings - Fork 108
/
Copy pathREADME.md
80 lines (60 loc) · 14.5 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
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
## Introduction
<a href="https://github.com/tensorflow/models/tree/master/research/deeplab">Official Repo</a>
<a href="https://github.com/SegmentationBLWX/sssegmentation/blob/main/ssseg/modules/models/backbones/mobilenet.py">Code Snippet</a>
<details>
<summary align="left"><a href="https://arxiv.org/pdf/1801.04381.pdf">MobileNetV2 (CVPR'2018)</a></summary>
```latex
@inproceedings{sandler2018mobilenetv2,
title={Mobilenetv2: Inverted residuals and linear bottlenecks},
author={Sandler, Mark and Howard, Andrew and Zhu, Menglong and Zhmoginov, Andrey and Chen, Liang-Chieh},
booktitle={Proceedings of the IEEE conference on computer vision and pattern recognition},
pages={4510--4520},
year={2018}
}
```
</details>
<details>
<summary align="left"><a href="https://arxiv.org/pdf/1905.02244.pdf">MobileNetV3 (ICCV'2019)</a></summary>
```latex
@inproceedings{Howard_2019_ICCV,
title={Searching for MobileNetV3},
author={Howard, Andrew and Sandler, Mark and Chu, Grace and Chen, Liang-Chieh and Chen, Bo and Tan, Mingxing and Wang, Weijun and Zhu, Yukun and Pang, Ruoming and Vasudevan, Vijay and Le, Quoc V. and Adam, Hartwig},
booktitle={The IEEE International Conference on Computer Vision (ICCV)},
pages={1314-1324},
month={October},
year={2019},
doi={10.1109/ICCV.2019.00140}}
}
```
</details>
## Results
#### PASCAL VOC
| Segmentor | Pretrain | Backbone | Crop Size | Schedule | Train/Eval Set | mIoU | Download |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
| FCN | ImageNet-1k-224x224 | M-V2-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/60 | trainaug/val | 59.89% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/fcn_mobilenetv2os8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/fcn_mobilenetv2os8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/fcn_mobilenetv2os8_voc.log) |
| PSPNet | ImageNet-1k-224x224 | M-V2-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/60 | trainaug/val | 68.40% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/pspnet_mobilenetv2os8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/pspnet_mobilenetv2os8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/pspnet_mobilenetv2os8_voc.log) |
| DeepLabV3 | ImageNet-1k-224x224 | M-V2-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/60 | trainaug/val | 70.08% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/deeplabv3_mobilenetv2os8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3_mobilenetv2os8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3_mobilenetv2os8_voc.log) |
| DeepLabV3plus | ImageNet-1k-224x224 | M-V2-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/60 | trainaug/val | 70.04% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/deeplabv3plus_mobilenetv2os8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3plus_mobilenetv2os8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3plus_mobilenetv2os8_voc.log) |
| LRASPPNet | - | M-V3S-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/180 | trainaug/val | 62.13% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/lrasppnet/lrasppnet_mobilenetv3sos8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3sos8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3sos8_voc.log) |
| LRASPPNet | - | M-V3L-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/180 | trainaug/val | 67.90% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/lrasppnet/lrasppnet_mobilenetv3los8_voc.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3los8_voc.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3los8_voc.log) |
#### ADE20k
| Segmentor | Pretrain | Backbone | Crop Size | Schedule | Train/Eval Set | mIoU | Download |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
| FCN | ImageNet-1k-224x224 | M-V2-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/130 | train/val | 30.85% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/fcn_mobilenetv2os8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/fcn_mobilenetv2os8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/fcn_mobilenetv2os8_ade20k.log) |
| PSPNet | ImageNet-1k-224x224 | M-V2-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/130 | train/val | 35.09% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/pspnet_mobilenetv2os8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/pspnet_mobilenetv2os8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/pspnet_mobilenetv2os8_ade20k.log) |
| DeepLabV3 | ImageNet-1k-224x224 | M-V2-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/130 | train/val | 37.55% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/deeplabv3_mobilenetv2os8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3_mobilenetv2os8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3_mobilenetv2os8_ade20k.log) |
| DeepLabV3plus | ImageNet-1k-224x224 | M-V2-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/130 | train/val | 37.66% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/deeplabv3plus_mobilenetv2os8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3plus_mobilenetv2os8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3plus_mobilenetv2os8_ade20k.log) |
| LRASPPNet | - | M-V3S-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/390 | train/val | 26.09% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/lrasppnet/lrasppnet_mobilenetv3sos8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3sos8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3sos8_ade20k.log) |
| LRASPPNet | - | M-V3L-D8 | 512x512 | LR/POLICY/BS/EPOCH: 0.01/poly/16/390 | train/val | 30.06% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/lrasppnet/lrasppnet_mobilenetv3los8_ade20k.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3los8_ade20k.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3los8_ade20k.log) |
#### CityScapes
| Segmentor | Pretrain | Backbone | Crop Size | Schedule | Train/Eval Set | mIoU | Download |
| :-: | :-: | :-: | :-: | :-: | :-: | :-: | :-: |
| FCN | ImageNet-1k-224x224 | M-V2-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 70.77% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/fcn_mobilenetv2os8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/fcn_mobilenetv2os8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/fcn_mobilenetv2os8_cityscapes.log) |
| PSPNet | ImageNet-1k-224x224 | M-V2-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 73.64% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/pspnet_mobilenetv2os8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/pspnet_mobilenetv2os8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/pspnet_mobilenetv2os8_cityscapes.log) |
| DeepLabV3 | ImageNet-1k-224x224 | M-V2-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 76.74% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/deeplabv3_mobilenetv2os8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3_mobilenetv2os8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3_mobilenetv2os8_cityscapes.log) |
| DeepLabV3plus | ImageNet-1k-224x224 | M-V2-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/8/220 | train/val | 76.68% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/mobilenetv2/deeplabv3plus_mobilenetv2os8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3plus_mobilenetv2os8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/deeplabv3plus_mobilenetv2os8_cityscapes.log) |
| LRASPPNet | - | M-V3S-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/16/660 | train/val | 65.06% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/lrasppnet/lrasppnet_mobilenetv3sos8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3sos8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3sos8_cityscapes.log) |
| LRASPPNet | - | M-V3L-D8 | 512x1024 | LR/POLICY/BS/EPOCH: 0.01/poly/16/660 | train/val | 69.98% | [cfg](https://raw.githubusercontent.com/SegmentationBLWX/sssegmentation/main/ssseg/configs/lrasppnet/lrasppnet_mobilenetv3los8_cityscapes.py) | [model](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3los8_cityscapes.pth) | [log](https://github.com/SegmentationBLWX/modelstore/releases/download/ssseg_mobilenet/lrasppnet_mobilenetv3los8_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**