[] (#3205)

Signed-off-by: Yulv-git <[email protected]>

Author: Yulv-git

EISeg/docs/remote_sensing.md

@@ -41,7 +41,7 @@ conda install gdal
 
 ### 2.1 数据加载
 
-目前EISeg仅支持了*.tif/tiff图像后缀的的遥感影像读取，由于训练数据都是来自于RGB三通道的遥感图像切片，因此交互分割也仅在RGB三通道上完成，也就表示EISeg支持多波段数据的波段选择。
+目前EISeg仅支持了*.tif/tiff图像后缀的遥感影像读取，由于训练数据都是来自于RGB三通道的遥感图像切片，因此交互分割也仅在RGB三通道上完成，也就表示EISeg支持多波段数据的波段选择。
 
 当使用EISeg打开GTiff图像时，会获取当前波段数，可通过波段设置的下拉列表进行设置。默认为[b1, b1, b1]。下例展示的是天宫一号多光谱数据设置真彩色：
 
@@ -63,7 +63,7 @@ conda install gdal
 当打开标注的GTiff图像带有地理参考，可设置EISeg保存时保存为带有地理参考的GTiff和ESRI Shapefile。
 
 - GTiff：已成为GIS和卫星遥感应用的行业图像标准文件。
-- ESRI Shapefile：是最常见的的矢量数据格式，Shapefile文件是美国环境系统研究所（ESRI）所研制的GIS文件系统格式文件，是工业标准的矢量数据文件。 所有的商业和开源GIS软件都支持。无处不在的它已成为行业标准。
+- ESRI Shapefile：是最常见的矢量数据格式，Shapefile文件是美国环境系统研究所（ESRI）所研制的GIS文件系统格式文件，是工业标准的矢量数据文件。 所有的商业和开源GIS软件都支持。无处不在的它已成为行业标准。
 
 ![82jlu-no59o](https://user-images.githubusercontent.com/71769312/141137726-76457454-5e9c-4ad0-85d6-d03f658ee63c.gif)
 

Matting/deploy/human_matting_android_demo/README.md

@@ -63,7 +63,7 @@ This demo uses the MODNet with HRNet_W18 backbone to perform human matting. Plea
 
 In order to be able to infer on Android phones, the dynamic graph model needs to be exported as a static graph model, and the input size of the image should be fixed when exporting.
 
-First, update the [PaddleSeg](https://github.com/paddlepaddle/paddleseg/tree/develop) repository. Then `cd` to the `PaddleSeg/contrib/Matting` directory. Then put the downloaded modnet-hrnet_w18.pdparams (traing by youself is ok) on current directory（`PaddleSeg/contrib/Matting`). After that, fix the config file `configs/modnet_mobilenetv2.yml`(note: hrnet18 is used, but the config file `modnet_hrnet_w18.yml` is based on `modnet_mobilenetv2.yml`), where,modify the val_dataset field as follows:
+First, update the [PaddleSeg](https://github.com/paddlepaddle/paddleseg/tree/develop) repository. Then `cd` to the `PaddleSeg/contrib/Matting` directory. Then put the downloaded modnet-hrnet_w18.pdparams (training by youself is ok) on current directory（`PaddleSeg/contrib/Matting`). After that, fix the config file `configs/modnet_mobilenetv2.yml`(note: hrnet18 is used, but the config file `modnet_hrnet_w18.yml` is based on `modnet_mobilenetv2.yml`), where,modify the val_dataset field as follows:
 
 ``` yml
 val_dataset:

Matting/ppmatting/models/backbone/resnet_vd.py

@@ -272,7 +272,7 @@ def __init__(self,
                         block] if dilation_dict and block in dilation_dict else 1
 
                     # Actually block here is 'stage', and i is 'block' in 'stage'
-                    # At the stage 4, expand the the dilation_rate if given multi_grid
+                    # At the stage 4, expand the dilation_rate if given multi_grid
                     if block == 3:
                         dilation_rate = dilation_rate * multi_grid[i]
                     ###############################################################################

configs/pssl/README.md

@@ -24,7 +24,7 @@ Here we show the configuration files of two lightweight models, [STDC2](https://
 
 There is no need to do this step if [STDC2](https://paddleseg.bj.bcebos.com/dygraph/pssl/stdc2_pssl_pretrained/model.pdparams) and [PPLite-Seg-B](https://paddleseg.bj.bcebos.com/dygraph/pssl/pp_liteseg_stdc2_pssl_pretrained/model.pdparams) are used.
 
-Otherwise, checking the following steps for preparing the enviroment and the datasets:
+Otherwise, checking the following steps for preparing the environment and the datasets:
 
 1. Make sure that PaddleSeg is well installed. See the [installation guide](https://github.com/PaddlePaddle/PaddleSeg/blob/release/2.6/docs/install.md) for details.
 

contrib/DomainAdaptation/models/gscnn.py

@@ -154,7 +154,7 @@ def forward(self, x, feat_list, s_input):
             input_shape[2:],
             mode='bilinear',
             align_corners=self.align_corners)
-        cs = F.sigmoid(cs)  # Ouput of shape stream
+        cs = F.sigmoid(cs)  # Output of shape stream
 
         return [cs, ]
 

contrib/LaneSeg/README.md

@@ -217,7 +217,7 @@ python export.py --help
 python deploy/python/infer.py \
     --config output/export/deploy.yaml \
     --image_path data/test_images/3.jpg \
-    --save_dir ouput/results
+    --save_dir output/results
 ```
 
 Run the following command to view more parameters.

contrib/LaneSeg/README_CN.md

@@ -220,7 +220,7 @@ python export.py --help
 python deploy/python/infer.py \
     --config output/export/deploy.yaml \
     --image_path  data/test_images/3.jpg \
-    --save_dir ouput/results
+    --save_dir output/results
 ```
 
 更多参数信息请运行如下命令进行查看:

contrib/MedicalSeg/medicalseg/utils/utils.py

@@ -149,7 +149,7 @@ def get_image_list(image_path, valid_suffix=None, filter_key=None):
     Args:
     image_path(str): the image or image folder where you want to get a image list from.
     valid_suffix(tuple): Contain only the suffix you want to include.
-    filter_key(dict): the key(ignore case) and whether you want to include it. e.g.:{"segmentation": True} will futher filter the imagename with segmentation in it.
+    filter_key(dict): the key(ignore case) and whether you want to include it. e.g.:{"segmentation": True} will further filter the imagename with segmentation in it.
 
     """
     if valid_suffix is None:
@@ -213,18 +213,18 @@ def save_array(save_path, save_content, form, image_infor):
     """
     if not isinstance(save_content, dict):
         raise TypeError(
-            'The save_content need to be dict which the key is the save name and the value is the numpy array to be saved, but recieved {}'
+            'The save_content need to be dict which the key is the save name and the value is the numpy array to be saved, but received {}'
             .format(type(save_content)))
 
     for (key, val) in save_content.items():
         if not isinstance(val, np.ndarray):
-            raise TypeError('We only save numpy array, but recieved {}'.format(
+            raise TypeError('We only save numpy array, but received {}'.format(
                 type(val)))
         if len(val.shape) > 3:
             save_content[key] = np.squeeze(val)
 
     if not isinstance(form, Iterable):
-        raise TypeError('The form need be iterable, but recieved {}'.format(
+        raise TypeError('The form need be iterable, but received {}'.format(
             type(form)))
 
     if save_path is not None:

contrib/MedicalSeg/nnunet/datasets/dataset.py

@@ -343,7 +343,7 @@ def setup_DA_params(self):
         if self.stage == 1 and self.cascade:
             self.data_aug_params["num_cached_per_thread"] = 2
 
-            self.data_aug_params['move_last_seg_chanel_to_data'] = True
+            self.data_aug_params['move_last_seg_channel_to_data'] = True
             self.data_aug_params['cascade_do_cascade_augmentations'] = True
 
             self.data_aug_params['cascade_random_binary_transform_p'] = 0.4

contrib/MedicalSeg/nnunet/transforms/augmentation.py

@@ -150,8 +150,8 @@ def get_moreDA_augmentation(dataloader_train,
 
     tr_transforms.append(RemoveLabelTransform(-1, 0))
 
-    if params.get("move_last_seg_chanel_to_data") is not None and params.get(
-            "move_last_seg_chanel_to_data"):
+    if params.get("move_last_seg_channel_to_data") is not None and params.get(
+            "move_last_seg_channel_to_data"):
         tr_transforms.append(
             MoveSegAsOneHotToData(
                 1, params.get("all_segmentation_labels"), 'seg', 'data'))
@@ -224,8 +224,8 @@ def get_moreDA_augmentation(dataloader_train,
         val_transforms.append(
             SegChannelSelectionTransform(params.get("selected_seg_channels")))
 
-    if params.get("move_last_seg_chanel_to_data") is not None and params.get(
-            "move_last_seg_chanel_to_data"):
+    if params.get("move_last_seg_channel_to_data") is not None and params.get(
+            "move_last_seg_channel_to_data"):
         val_transforms.append(
             MoveSegAsOneHotToData(
                 1, params.get("all_segmentation_labels"), 'seg', 'data'))

contrib/MedicalSeg/nnunet/transforms/default_config.py

@@ -49,7 +49,7 @@
     "mask_was_used_for_normalization": None,
     "border_mode_data": "constant",
     "all_segmentation_labels": None,
-    "move_last_seg_chanel_to_data": False,
+    "move_last_seg_channel_to_data": False,
     "cascade_do_cascade_augmentations": False,
     "cascade_random_binary_transform_p": 0.4,
     "cascade_random_binary_transform_p_per_label": 1,

contrib/PP-HumanSeg/README.md

@@ -57,7 +57,7 @@ PP-HumanSegV2-Lite protrait segmentation model: **The inference speed is increas
   * Faster inference speed: We reduce the input resolution, which reduces the inference time and increases the receptive field.
   * Better robustness: Based on the idea of transfer learning, we first pretrain the model on a large general human segmentation dataset, and then finetune it on a small portrait segmentation dataset.
 
-| Model Name | Best Input Shape | mIou(%) | Inference Time on Arm CPU(ms) | Modle Size(MB) | Config File | Links |
+| Model Name | Best Input Shape | mIou(%) | Inference Time on Arm CPU(ms) | Model Size(MB) | Config File | Links |
 | --- | --- | --- | ---| --- | --- | --- |
 | PP-HumanSegV1-Lite | 398x224 | 93.60 | 29.68 | 2.3 | [cfg](./configs/portrait_pp_humansegv1_lite.yml) | [Checkpoint](https://paddleseg.bj.bcebos.com/dygraph/pp_humanseg_v2/portrait_pp_humansegv1_lite_398x224_pretrained.zip) \| [Inference Model (Argmax)](https://paddleseg.bj.bcebos.com/dygraph/pp_humanseg_v2/portrait_pp_humansegv1_lite_398x224_inference_model.zip) \| [Inference Model (Softmax)](https://paddleseg.bj.bcebos.com/dygraph/pp_humanseg_v2/portrait_pp_humansegv1_lite_398x224_inference_model_with_softmax.zip) |
 | PP-HumanSegV2-Lite | 256x144 | 96.63 | 15.86 | 5.4 | [cfg](./configs/portrait_pp_humansegv2_lite.yml) | [Checkpoint](https://paddleseg.bj.bcebos.com/dygraph/pp_humanseg_v2/portrait_pp_humansegv2_lite_256x144_smaller/portrait_pp_humansegv2_lite_256x144_pretrained.zip) \| [Inference Model (Argmax)](https://paddleseg.bj.bcebos.com/dygraph/pp_humanseg_v2/portrait_pp_humansegv2_lite_256x144_smaller/portrait_pp_humansegv2_lite_256x144_inference_model.zip) \| [Inference Model (Softmax)](https://paddleseg.bj.bcebos.com/dygraph/pp_humanseg_v2/portrait_pp_humansegv2_lite_256x144_smaller/portrait_pp_humansegv2_lite_256x144_inference_model_with_softmax.zip) |

contrib/PaddleLabel/doc/CN/project/instance_segmentation.md

@@ -120,7 +120,7 @@ category{
 ### 掩膜标注
 
 1. 点击右侧“添加标签”，填写信息并创建标签
-1. 选择一个标签，点击左侧工具栏的“笔刷”（鼠标悬浮可以修改笔刷大小），在图像界面上按住鼠标左键绘制需要标注的物体内部，实例分割可以反复选择同一标签标注不同的实例，，需要删除掩膜可以点击左侧工具栏的“橡皮擦”进行修改
+1. 选择一个标签，点击左侧工具栏的“笔刷”（鼠标悬浮可以修改笔刷大小），在图像界面上按住鼠标左键绘制需要标注的物体内部，实例分割可以反复选择同一标签标注不同的实例，需要删除掩膜可以点击左侧工具栏的“橡皮擦”进行修改
 1. **标注完成一个实例后，点击右上角"确定轮廓"进行实例确认**
 1. 点击左右按钮切换图像，重复上述操作，直到所有数据标注完毕
 1. 下方进度展示可以查看标注进度

contrib/PaddleLabel/doc/CN/training/PdLabel_PdClas.md

@@ -78,7 +78,7 @@ Arch:
   name: ShuffleNetV2_x0_25
   class_num: 3
 
-# loss function config for traing/eval process
+# loss function config for training/eval process
 Loss:
   Train:
     - CELoss:

contrib/PaddleLabel/doc/EN/README.md

@@ -8,7 +8,7 @@ PaddleLabel aims to become an effective and flexible data annotation tool. There
 
 ## Install
 
-Installing in a new enviroment is not required but suggested.
+Installing in a new environment is not required but suggested.
 
 ```python
 conda create -n pplabel python=3.9

contrib/PanopticDeepLab/README.md

@@ -4,7 +4,7 @@ English | [简体中文](README_CN.md)
 
 The implementation of [Panoptic Deeplab](https://arxiv.org/abs/1911.10194) algorithm based on PaddlePaddle.
 
-Panoptic DeepLab has comfirmed that the bottem-up algorithm can achieve state-of-the-art effect for the first time。Panoptic DeepLab predicts three ouputs: Semantic Segmentation, Center Prediction and Center Regression。According to the principle of nearest distance, the pixels of instance category are gathered to the center of the instance to obtain the segmentation result. Finally, according to the rules of majority-vote, semantic segmentation results and instance segmentation results are fused to obtain the final panoptic segmentation results.
+Panoptic DeepLab has comfirmed that the bottem-up algorithm can achieve state-of-the-art effect for the first time。Panoptic DeepLab predicts three outputs: Semantic Segmentation, Center Prediction and Center Regression。According to the principle of nearest distance, the pixels of instance category are gathered to the center of the instance to obtain the segmentation result. Finally, according to the rules of majority-vote, semantic segmentation results and instance segmentation results are fused to obtain the final panoptic segmentation results.
 It achieves segmentation by assigning each pixel to one category or instance.
 ![](./docs/panoptic_deeplab.jpg)
 

contrib/PanopticSeg/paddlepanseg/runners/maskformer_runner.py

@@ -432,7 +432,7 @@ def compute_losses(self, net_out, data):
         # Retrieve the matching between the outputs and the targets
         indices = self.matcher(data, net_out)
 
-        # Compute the average number of target boxes accross all nodes, for normalization purposes
+        # Compute the average number of target boxes across all nodes, for normalization purposes
         num_masks = sum(len(gt_ids) for gt_ids in data['gt_ids'])
         num_masks = paddle.to_tensor([num_masks], dtype='float32')
         n_ranks = paddle.distributed.get_world_size()

deploy/fastdeploy/semantic_segmentation/android/README.md

@@ -1,7 +1,7 @@
 [English](README.md) | 简体中文
 # 图像分割 PaddleSeg Android Demo 使用文档  
 
-在 Android 上实现实时的人像分割功能，此 Demo 有很好的的易用性和开放性，如在 Demo 中跑自己训练好的模型等。
+在 Android 上实现实时的人像分割功能，此 Demo 有很好的易用性和开放性，如在 Demo 中跑自己训练好的模型等。
 
 ## 环境准备
 

deploy/fastdeploy/semantic_segmentation/rockchip/rv1126/README.md

@@ -29,7 +29,7 @@ PaddleSeg模型导出，请参考其文档说明[模型导出](https://github.co
 ### 3.3 导出须知  
 请参考[模型导出](https://github.com/PaddlePaddle/PaddleSeg/blob/develop/docs/model_export_cn.md)中`output_op`参数的说明，获取您部署所需的模型，比如是否带`argmax`或`softmax`算子
 
-### 3.4 转换为为RV1126支持的INT8模型
+### 3.4 转换为RV1126支持的INT8模型
 瑞芯微RV1126仅支持INT8，将推理模型量化压缩为INT8模型，FastDeploy模型量化的方法及一键自动化压缩工具可以参考[模型量化](../../../quantize/README.md)
 
 ## 4. 详细的部署示例

deploy/python/infer_dataset.py

@@ -309,7 +309,7 @@ def main(args):
 if __name__ == '__main__':
     """
     Based on the infer config and dataset, this program read the test and
-    label images, applys the transfors, run the predictor, ouput the accuracy.
+    label images, applys the transfors, run the predictor, output the accuracy.
 
     For example:
     python deploy/python/infer_dataset.py \

docs/deployment/inference/inference_cn.md

@@ -16,7 +16,7 @@
 
 在终端输入以下命令进行预测:
 ```shell
-python deploy/python/infer.py --config /path/to/deploy.yaml --image_path # 如果导出指定了input_shape，传入infer的的图片大小或者经过deploy.yaml变换的图片大小需要和 input_shape 保持一致
+python deploy/python/infer.py --config /path/to/deploy.yaml --image_path # 如果导出指定了input_shape，传入infer的图片大小或者经过deploy.yaml变换的图片大小需要和 input_shape 保持一致
 ```
 
 参数说明如下:

docs/deployment/slim/distill/distill.md

@@ -58,7 +58,7 @@ python tools/train.py \
     --save_dir output/deeplabv3p_resnet50
 ```
 
-After the traing, the mIoU of the teacher model is 91.54% and the trained weights are saved in `output/deeplabv3p_resnet50/best_model/model.pdparams`.
+After the training, the mIoU of the teacher model is 91.54% and the trained weights are saved in `output/deeplabv3p_resnet50/best_model/model.pdparams`.
 
 ### 2.4 Train the Student Model (Optional)
 

docs/deployment/slim/prune/prune.md

@@ -1,6 +1,6 @@
 English | [简体中文](prune_cn.md)
 
-# Model Pruning Toturial
+# Model Pruning Tutorial
 
 Many CNN models require huge computing and memory overhead, which seriously hinders the applications under limited resources. Model compression can reduce model parameters or FLOPs, and facilitate the deployment of restricted hardware. Powered by PaddleSlim, PaddleSeg provides model pruning for developers in image segmentation.
 

docs/pr/pr/style_cn.md

@@ -324,7 +324,7 @@ class Cityscapes(Dataset):
 1. ```__init__```中参数全部显式写出，**不能**包括变长参数比如:```*args, **kwargs```；
 2. ```super().__init__()```保持空参数；
 3. 参数顺序和上述参数顺序一致；
-4. 通过在在```__init__```方法中建立 ```self.file_list```，之后就根据其中元素的路径读取对应图片。
+4. 通过在```__init__```方法中建立 ```self.file_list```，之后就根据其中元素的路径读取对应图片。
 
 ## 3 导出和测试预测模型
 

docs/whole_process.md

@@ -401,7 +401,7 @@ output
 | Python | Paddle prediction library | [e.g.](../deploy/python/) |
 | C++ | Paddle prediction library | [e.g.](../deploy/cpp/) |
 | Mobile | PaddleLite   | [e.g.](../deploy/lite/) |
-| Serving | HubServing   | Comming soon |
+| Serving | HubServing   | Coming soon |
 | Front-end | PaddleJS     | [e.g.](../deploy/web/) |
 
 ```

paddleseg/models/backbones/resnet_vd.py

@@ -297,7 +297,7 @@ def __init__(self,
                         block] if dilation_dict and block in dilation_dict else 1
 
                     # Actually block here is 'stage', and i is 'block' in 'stage'
-                    # At the stage 4, expand the the dilation_rate if given multi_grid
+                    # At the stage 4, expand the dilation_rate if given multi_grid
                     if block == 3:
                         dilation_rate = dilation_rate * multi_grid[i]
                     ###############################################################################

paddleseg/models/backbones/top_transformer.py

@@ -593,23 +593,23 @@ def init_weight(self):
             utils.load_entire_model(self, self.pretrained)
 
     def forward(self, x):
-        ouputs = self.tpm(x)
-        out = self.ppa(ouputs)
+        outputs = self.tpm(x)
+        out = self.ppa(outputs)
         out = self.trans(out)
 
         if self.injection:
             xx = out.split(self.feat_channels, axis=1)
             results = []
             for i in range(len(self.feat_channels)):
                 if i in self.trans_out_indices:
-                    local_tokens = ouputs[i]
+                    local_tokens = outputs[i]
                     global_semantics = xx[i]
                     out_ = self.SIM[i](local_tokens, global_semantics)
                     results.append(out_)
             return results
         else:
-            ouputs.append(out)
-            return ouputs
+            outputs.append(out)
+            return outputs
 
 
 @manager.BACKBONES.add_component