Skip to content

Commit 60eb626

Browse files
zh320zh320
committed
update README
Former-commit-id: d8b3b0f9e91446010b56ef6f71df71a148388b46 [formerly df9990a] Former-commit-id: 4484e2ec934881eb1a95092e446b8c4e93987bdd
1 parent 48430f0 commit 60eb626

File tree

1 file changed

+16
-16
lines changed

1 file changed

+16
-16
lines changed

‎README.md

Lines changed: 16 additions & 16 deletions
Original file line numberDiff line numberDiff line change
@@ -133,7 +133,7 @@ CUDA_VISIBLE_DEVICES=0,1,2,3 python main.py
133133
| ADSCNet | 2019 | None | n.a. / 0.51 | 89 | n.a. / 67.5 | [69.06](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/adscnet.pth) |
134134
| AGLNet | 2020 | None | 1.12 / 1.02 | 61 | 69.39 / 70.1 | [73.58](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/aglnet.pth) |
135135
| BiSeNetv1 | 2018 | ResNet18 | 49.0 / 13.32 | 88 | 74.8 / 74.7 | [74.91](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/bisenetv1.pth) |
136-
| BiSeNetv2 | 2020 | None | n.a. / 2.27 | 142 | 73.4 / 72.6 | [73.73<sup>3</sup>](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/bisenetv2-aux.pth) |
136+
| BiSeNetv2 | 2020 | None | n.a. / 2.27 | 142 | 73.4 / 72.6 | [73.73](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/bisenetv2-aux.pth)<sup>3</sup> |
137137
| CANet | 2019 | MobileNetv2 | 4.8 / 4.77 | 76 | 73.4 / 73.5 | [76.59](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.1/canet.pth) |
138138
| CFPNet | 2021 | None | 0.55 / 0.27 | 64 | n.a. / 70.1 | [70.08](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/cfpnet.pth) |
139139
| CGNet | 2018 | None | 0.41 / 0.24 | 157 | 59.7 / 64.8<sup>4</sup> | [67.25](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/cgnet.pth) |
@@ -160,34 +160,34 @@ CUDA_VISIBLE_DEVICES=0,1,2,3 python main.py
160160
| MiniNetv2 | 2020 | None | 0.5 / 0.51 | 86 | n.a. / 70.5 | [71.79](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/mininetv2.pth) |
161161
| PP-LiteSeg | 2022 | STDC1 | n.a. / 6.33 | 201 | 76.0 / 74.9 | [72.49](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/ppliteseg_stdc1.pth) |
162162
| PP-LiteSeg | 2022 | STDC2 | n.a. / 10.56 | 136 | 78.2 / 77.5 | [74.37](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/ppliteseg_stdc2.pth) |
163-
| RegSeg | 2021 | None | 3.34 / 3.37 | 104 | 78.5 / 78.3 | 74.28 |
163+
| RegSeg | 2021 | None | 3.34 / 3.37 | 104 | 78.5 / 78.3 | [74.28](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/regseg.pth) |
164164
| SegNet | 2015 | None | 29.46 / 29.48 | 14 | n.a. / 56.1 | [70.77](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/segnet.pth) |
165165
| ShelfNet | 2018 | ResNet18 | 23.5 / 16.04 | 110 | n.a. / 74.8 | [77.63](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/shelfnet.pth) |
166166
| SQNet | 2016 | SqueezeNet-1.1 | n.a. / 4.81 | 69 | n.a. / 59.8 | [69.55](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/sqnet.pth) |
167-
| STDC | 2021 | STDC1 | n.a. / 7.79 | 163 | 74.5 / 75.3 | 75.25<sup>6</sup> |
168-
| STDC | 2021 | STDC2 | n.a. / 11.82 | 119 | 77.0 / 76.8 | 76.78<sup>6</sup> |
167+
| STDC | 2021 | STDC1 | n.a. / 7.79 | 163 | 74.5 / 75.3 | [75.25](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/stdc1.pth)<sup>6</sup> |
168+
| STDC | 2021 | STDC2 | n.a. / 11.82 | 119 | 77.0 / 76.8 | [76.78](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/stdc2.pth)<sup>6</sup> |
169169
| SwiftNet | 2019 | ResNet18 | 11.8 / 11.95 | 141 | 75.4 / 75.5 | [75.43](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.0/swiftnet.pth) |
170170

171171
[<sup>1</sup>FPSs are evaluated on RTX 2080 at resolution 1024x512 using this [script](tools/test_speed.py)]
172172
[<sup>2</sup>These results are obtained by training 800 epochs with crop-size 1024x1024]
173173
[<sup>3</sup>These results are obtained by using auxiliary heads]
174174
[<sup>4</sup>This result is obtained by using deeper model, i.e. CGNet_M3N21]
175175
[<sup>5</sup>The original encoder of ICNet is ResNet50]
176-
[<sup>6</sup>In my experiments, detail loss does not improve the performances. However, use auxiliary heads do contribute to the improvements]
176+
[<sup>6</sup>In my experiments, detail loss does not improve the performances. However, using auxiliary heads does contribute to the improvements]
177177

178178
## SMP performance on Cityscapes
179179

180-
| Decoder | Params (M) | mIoU (200 epoch) | mIoU (800 epoch) |
181-
|:-------------:|:----------:|:----------------:|:----------------:|
182-
| DeepLabv3 | 15.90 | 75.22 | 77.16 |
183-
| DeepLabv3Plus | 12.33 | 73.97 | 75.90 |
184-
| FPN | 13.05 | 73.44 | 74.94 |
185-
| LinkNet | 11.66 | 71.17 | 73.19 |
186-
| MANet | 21.68 | 74.59 | 76.14 |
187-
| PAN | 11.37 | 70.25 | 72.46 |
188-
| PSPNet | 11.41 | 61.63 | 67.26 |
189-
| UNet | 14.33 | 72.99 | 74.45 |
190-
| UNetPlusPlus | 15.97 | 74.31 | 75.57 |
180+
| Decoder | Params (M) | mIoU (200 epoch) | mIoU (800 epoch) |
181+
|:-------------:|:----------:|:----------------:|:--------------------------------------------------------------------------------------------------------------:|
182+
| DeepLabv3 | 15.90 | 75.22 | [77.16](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/deeplabv3.pth) |
183+
| DeepLabv3Plus | 12.33 | 73.97 | [75.90](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/deeplabv3p.pth) |
184+
| FPN | 13.05 | 73.44 | [74.94](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/fpn.pth) |
185+
| LinkNet | 11.66 | 71.17 | [73.19](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/linknet.pth) |
186+
| MANet | 21.68 | 74.59 | [76.14](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/manet.pth) |
187+
| PAN | 11.37 | 70.25 | [72.46](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/pan.pth) |
188+
| PSPNet | 11.41 | 61.63 | [67.26](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/pspnet.pth) |
189+
| UNet | 14.33 | 72.99 | [74.45](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/unet.pth) |
190+
| UNetPlusPlus | 15.97 | 74.31 | [75.57](https://github.com/zh320/realtime-semantic-segmentation-pytorch/releases/download/v1.2/unetpp.pth) |
191191

192192
[For comparison, the above results are all using ResNet-18 as encoders.]
193193

0 commit comments

Comments
 (0)