Skip to content

Commit 83466c1

Browse files
Alexander Kirillovfacebook-github-bot
authored andcommitted
Add semantic segmenation to PointRend readme
Summary: Add Cityscapes model Reviewed By: ppwwyyxx Differential Revision: D21358667 fbshipit-source-id: 5c86f937eb0d8a9d1d89bbd49ce471c77934b272
1 parent 47c34bc commit 83466c1

File tree

1 file changed

+23
-3
lines changed

1 file changed

+23
-3
lines changed

projects/PointRend/README.md

Lines changed: 23 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,7 @@ Alexander Kirillov, Yuxin Wu, Kaiming He, Ross Girshick
88
<img src="https://alexander-kirillov.github.io/images/kirillov2019pointrend.jpg"/>
99
</div><br/>
1010

11-
In this repository, we release code for PointRend in Detectron2. PointRend can be flexibly applied to both instance and semantic (**comming soon**) segmentation tasks by building on top of existing state-of-the-art models.
11+
In this repository, we release code for PointRend in Detectron2. PointRend can be flexibly applied to both instance and semantic segmentation tasks by building on top of existing state-of-the-art models.
1212

1313
## Installation
1414
Install Detectron 2 following [INSTALL.md](https://github.com/facebookresearch/detectron2/blob/master/INSTALL.md). You are ready to go!
@@ -86,7 +86,7 @@ Cityscapes model is trained with ImageNet pretraining.
8686
<th valign="bottom">model id</th>
8787
<th valign="bottom">download</th>
8888
<!-- TABLE BODY -->
89-
<tr><td align="left"><a href="configs/InstanceSegmentation/pointrend_rcnn_R_50_FPN_1x_cs.yaml">PointRend</a></td>
89+
<tr><td align="left"><a href="configs/InstanceSegmentation/pointrend_rcnn_R_50_FPN_1x_cityscapes.yaml">PointRend</a></td>
9090
<td align="center">R50-FPN</td>
9191
<td align="center">1&times;</td>
9292
<td align="center">224&times;224</td>
@@ -99,7 +99,27 @@ Cityscapes model is trained with ImageNet pretraining.
9999

100100
## Semantic Segmentation
101101

102-
**[comming soon]**
102+
#### Cityscapes
103+
Cityscapes model is trained with ImageNet pretraining.
104+
105+
<table><tbody>
106+
<!-- START TABLE -->
107+
<!-- TABLE HEADER -->
108+
<th valign="bottom">Method</th>
109+
<th valign="bottom">Backbone</th>
110+
<th valign="bottom">Output<br/>resolution</th>
111+
<th valign="bottom">mIoU</th>
112+
<th valign="bottom">model id</th>
113+
<th valign="bottom">download</th>
114+
<!-- TABLE BODY -->
115+
<tr><td align="left"><a href="configs/SemanticSegmentation/pointrend_semantic_R_50_FPN_1x_cityscapes.yaml">SemanticFPN + PointRend</a></td>
116+
<td align="center">R50-FPN</td>
117+
<td align="center">1024&times;2048</td>
118+
<td align="center">78.6</td>
119+
<td align="center">186480235</td>
120+
<td align="center"><a href="https://dl.fbaipublicfiles.com/detectron2/PointRend/SemanticSegmentation/pointrend_semantic_R_50_FPN_1x_cityscapes/186480235/model_final_5f3665.pkl">model</a>&nbsp;|&nbsp;<a href="https://dl.fbaipublicfiles.com/detectron2/PointRend/SemanticSegmentation/pointrend_semantic_R_50_FPN_1x_cityscapes/186480235/metrics.json">metrics</a></td>
121+
</tr>
122+
</tbody></table>
103123

104124
## <a name="CitingPointRend"></a>Citing PointRend
105125

0 commit comments

Comments
 (0)