PaddlePaddle · LutaoChu · Aug 4, 2021 · Jul 27, 2021 · Jul 27, 2021 · Jul 27, 2021
diff --git a/configs/README.md b/configs/README.md
@@ -2,29 +2,29 @@
 
 |模型\骨干网络|ResNet50|ResNet101|HRNetw18|HRNetw48|
 |-|-|-|-|-|
-|[ANN](./configs/ann)|✔|✔|||
-|[BiSeNetv2](./configs/bisenet)|-|-|-|-|
-|[DANet](./configs/danet)|✔|✔|||
-|[Deeplabv3](./configs/deeplabv3)|✔|✔|||
-|[Deeplabv3P](./configs/deeplabv3p)|✔|✔|||
-|[Fast-SCNN](./configs/fastscnn)|-|-|-|-|
-|[FCN](./configs/fcn)|||✔|✔|
-|[GCNet](./configs/gcnet)|✔|✔|||
-|[GSCNN](./configs/gscnn)|✔|✔|||
-|[HarDNet](./configs/hardnet)|-|-|-|-|
-|[OCRNet](./configs/ocrnet/)|||✔|✔|
-|[PSPNet](./configs/pspnet)|✔|✔|||
-|[U-Net](./configs/unet)|-|-|-|-|
-|[U<sup>2</sup>-Net](./configs/u2net)|-|-|-|-|
-|[Att U-Net](./configs/attention_unet)|-|-|-|-|
-|[U-Net++](./configs/unet_plusplus)|-|-|-|-|
-|[U-Net3+](./configs/unet_3plus)|-|-|-|-|
-|[DecoupledSegNet](./configs/decoupled_segnet)|✔|✔|||
-|[EMANet](./configs/emanet)|✔|✔|-|-|
-|[ISANet](./configs/isanet)|✔|✔|-|-|
-|[DNLNet](./configs/dnlnet)|✔|✔|-|-|
-|[SFNet](./configs/sfnet)|✔|-|-|-|
-|[PPSegLite](./configs/ppseg_lite)|-|-|-|-|
+|[ANN](./ann)|✔|✔|||
+|[BiSeNetv2](./bisenet)|-|-|-|-|
+|[DANet](./danet)|✔|✔|||
+|[Deeplabv3](./deeplabv3)|✔|✔|||
+|[Deeplabv3P](./deeplabv3p)|✔|✔|||
+|[Fast-SCNN](./fastscnn)|-|-|-|-|
+|[FCN](./fcn)|||✔|✔|
+|[GCNet](./gcnet)|✔|✔|||
+|[GSCNN](./gscnn)|✔|✔|||
+|[HarDNet](./hardnet)|-|-|-|-|
+|[OCRNet](./ocrnet/)|||✔|✔|
+|[PSPNet](./pspnet)|✔|✔|||
+|[U-Net](./unet)|-|-|-|-|
+|[U<sup>2</sup>-Net](./u2net)|-|-|-|-|
+|[Att U-Net](./attention_unet)|-|-|-|-|
+|[U-Net++](./unet_plusplus)|-|-|-|-|
+|[U-Net3+](./unet_3plus)|-|-|-|-|
+|[DecoupledSegNet](./decoupled_segnet)|✔|✔|||
+|[EMANet](./emanet)|✔|✔|-|-|
+|[ISANet](./isanet)|✔|✔|-|-|
+|[DNLNet](./dnlnet)|✔|✔|-|-|
+|[SFNet](./sfnet)|✔|-|-|-|
+|[PPSegLite](./ppseg_lite)|-|-|-|-|
 
 # 模型说明
 

diff --git a/docs/annotation/cityscapes_demo/cityscapes_demo_dataset.yaml b/docs/annotation/cityscapes_demo/cityscapes_demo_dataset.yaml
diff --git a/...es_demo/gtFine/train/stuttgart/stuttgart_000021_000019_gtFine_labelTrainIds.png b/...es_demo/gtFine/train/stuttgart/stuttgart_000021_000019_gtFine_labelTrainIds.png
diff --git a/...es_demo/gtFine/train/stuttgart/stuttgart_000072_000019_gtFine_labelTrainIds.png b/...es_demo/gtFine/train/stuttgart/stuttgart_000072_000019_gtFine_labelTrainIds.png
diff --git a/...apes_demo/gtFine/val/frankfurt/frankfurt_000001_062250_gtFine_labelTrainIds.png b/...apes_demo/gtFine/val/frankfurt/frankfurt_000001_062250_gtFine_labelTrainIds.png
diff --git a/...apes_demo/gtFine/val/frankfurt/frankfurt_000001_063045_gtFine_labelTrainIds.png b/...apes_demo/gtFine/val/frankfurt/frankfurt_000001_063045_gtFine_labelTrainIds.png
diff --git a/...scapes_demo/leftImg8bit/train/stuttgart/stuttgart_000021_000019_leftImg8bit.png b/...scapes_demo/leftImg8bit/train/stuttgart/stuttgart_000021_000019_leftImg8bit.png
diff --git a/...scapes_demo/leftImg8bit/train/stuttgart/stuttgart_000072_000019_leftImg8bit.png b/...scapes_demo/leftImg8bit/train/stuttgart/stuttgart_000072_000019_leftImg8bit.png
diff --git a/...tyscapes_demo/leftImg8bit/val/frankfurt/frankfurt_000001_062250_leftImg8bit.png b/...tyscapes_demo/leftImg8bit/val/frankfurt/frankfurt_000001_062250_leftImg8bit.png
diff --git a/...tyscapes_demo/leftImg8bit/val/frankfurt/frankfurt_000001_063045_leftImg8bit.png b/...tyscapes_demo/leftImg8bit/val/frankfurt/frankfurt_000001_063045_leftImg8bit.png
diff --git a/docs/annotation/cityscapes_demo/train_list.txt b/docs/annotation/cityscapes_demo/train_list.txt
diff --git a/docs/annotation/cityscapes_demo/val_list.txt b/docs/annotation/cityscapes_demo/val_list.txt
diff --git a/docs/data/marker/marker.md b/docs/data/marker/marker.md
@@ -48,8 +48,8 @@ We want to write the path of the image to the three folders `train.txt`, `val.tx
 The texts of `train.txt`, `val.txt` and `test.txt` are divided into two columns with spaces as separators. The first column is the relative path of the image file relative to the dataset, and the second column is the relative path of the image file The relative path of the dataset. As follows:
 
 ```
-images/xxx1.jpg (xx1.png) annotations/xxx1.png
-images/xxx2.jpg (xx2.png) annotations/xxx2.png
+images/xxx1.jpg  annotations/xxx1.png
+images/xxx2.jpg  annotations/xxx2.png
 ...
 ```
 `labels.txt`: Each row has a separate category, and the corresponding row number is the id corresponding to the category (the row number starts from 0), as shown below:
@@ -94,7 +94,7 @@ The data file structure is as follows:
 ```
 ./dataset/  # Dataset root directory
 |--images  # Original image catalog
-|  |--xxx1.jpg (xx1.png)
+|  |--xxx1.jpg
 |  |--...
 |  └--...
 |
@@ -146,6 +146,7 @@ python tools/split_dataset_list.py <dataset_root> images annotations --split 0.6
 PaddleSeg uses a common file list method to organize training set, validation set and test set. The corresponding file list must be prepared before the training, evaluation, and visualization process.
 
 It is recommended to organize it into the following structure:
+
     custom_dataset
         |
         |--images
@@ -166,8 +167,8 @@ It is recommended to organize it into the following structure:
 
 The contents of train.txt and val.txt are as follows:
 
-    images/image1.jpg labels/label1.png
-    images/image2.jpg labels/label2.png
+    images/image1.jpg
+    images/image2.jpg
     ...
 
 **NOTE**
@@ -183,16 +184,8 @@ If the dataset lacks annotated pictures, the file list does not need to include
 **NOTE**
 
 The file list at this time can only be used when calling `predict.py` for visual display.
-Please use spaces as the delimiter, that is, the contents of train.txt and val.txt are organized as follows:
-
-    images/image1.jpg labels/label1.png
-    images/image2.jpg labels/label2.png
-    ...
-
-
-**What is the list of compliant files?**
+You need to ensure that the separator of the file list is consistent with your Dataset class. The default separator is a space.
 
-Please use spaces as the separator (to be consistent with the [Dataset](../../../paddleseg/datasets/dataset.py) class), that is, the contents of train.txt and val.txt are organized as:
 
 ### 4.2 Organize the dataset directory structure
 

diff --git a/docs/data/marker/marker_cn.md b/docs/data/marker/marker_cn.md
@@ -46,8 +46,8 @@ python tools/gray2pseudo_color.py <dir_or_file> <output_dir> --dataset_dir <data
 我们希望将图像的路径写入到`train.txt`，`val.txt`，`test.txt`和`labels.txt`三个文件夹中，因为PaddleSeg是通过读取这些文本文件来定位图像路径的。
 `train.txt`，`val.txt`和`test.txt`文本以空格为分割符分为两列，第一列为图像文件相对于dataset的相对路径，第二列为标注图像文件相对于dataset的相对路径。如下所示：
 ```
-images/xxx1.jpg (xx1.png) annotations/xxx1.png
-images/xxx2.jpg (xx2.png) annotations/xxx2.png
+images/xxx1.jpg  annotations/xxx1.png
+images/xxx2.jpg  annotations/xxx2.png
 ...
 ```
 `labels.txt`: 每一行为一个单独的类别，相应的行号即为类别对应的id（行号从0开始)，如下所示：
@@ -93,7 +93,7 @@ PddleSeg已支持2种标注工具：LabelMe、EISeg交互式分割标注工具
 ```
 ./dataset/  # 数据集根目录
 |--images  # 原图目录
-|  |--xxx1.jpg (xx1.png)
+|  |--xxx1.jpg
 |  |--...
 |  └--...
 |
@@ -143,6 +143,7 @@ python tools/split_dataset_list.py <dataset_root> images annotations --split 0.6
 PaddleSeg采用通用的文件列表方式组织训练集、验证集和测试集。在训练、评估、可视化过程前必须准备好相应的文件列表。
 
 推荐整理成如下结构：
+
     custom_dataset
         |
         |--images
@@ -163,8 +164,8 @@ PaddleSeg采用通用的文件列表方式组织训练集、验证集和测试
 
 其中train.txt和val.txt的内容如下所示：
 
-    images/image1.jpg labels/label1.png
-    images/image2.jpg labels/label2.png
+    images/image1.jpg
+    images/image2.jpg
     ...
 
 
@@ -181,16 +182,8 @@ PaddleSeg采用通用的文件列表方式组织训练集、验证集和测试
 **注意事项**
 
 此时的文件列表仅可在调用`predict.py`进行可视化展示时使用，
-分割符请使用空格（与[Dataset](../../../paddleseg/datasets/dataset.py)类中保持一致），即train.txt和val.txt的内容整理为：
-
-    images/image1.jpg labels/label1.png
-    images/image2.jpg labels/label2.png
-    ...
-
-
-**符合规范的文件列表是什么样的呢？**
+需要保证文件列表的分割符与你的Dataset类保持一致，默认分割符为空格。
 
-请参考[目录](../../annotation/cityscapes_demo)。
 
 ### 4.2 数据集目录结构整理