提交 cb7fabbe 编写于 作者: L LDOUBLEV

fix comments

上级 693e4db5
...@@ -10,12 +10,13 @@ ...@@ -10,12 +10,13 @@
1. jeston安装PaddlePaddle 1. jeston安装PaddlePaddle
paddlepaddle下载[链接](https://www.paddlepaddle.org.cn/inference/user_guides/download_lib.html#python) PaddlePaddle下载[链接](https://www.paddlepaddle.org.cn/inference/user_guides/download_lib.html#python)
请选择适合的您Jetpack版本、cuda版本、trt版本的安装包。 请选择适合的您Jetpack版本、cuda版本、trt版本的安装包。
安装命令: 安装命令:
```shell ```shell
pip3 install -U paddlepaddle_gpu-*-cp36-cp36m-linux_aarch64.whl # 安装paddle,以paddlepaddle_gpu-2.3.0rc0-cp36-cp36m-linux_aarch64.whl 为例
pip3 install -U paddlepaddle_gpu-2.3.0rc0-cp36-cp36m-linux_aarch64.whl
``` ```
...@@ -37,7 +38,7 @@ pip3 install -r requirements.txt ...@@ -37,7 +38,7 @@ pip3 install -r requirements.txt
## 2. 执行预测 ## 2. 执行预测
[文档](../../doc/doc_ch/ppocr_introduction.md) 模型库中获取PPOCR模型,下面以PP-OCRv3模型为例,介绍在PPOCR模型在jeston上的使用方式: [文档](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/ppocr_introduction.md#6-%E6%A8%A1%E5%9E%8B%E5%BA%93) 模型库中获取PPOCR模型,下面以PP-OCRv3模型为例,介绍在PPOCR模型在jeston上的使用方式:
下载并解压PP-OCRv3模型 下载并解压PP-OCRv3模型
``` ```
......
...@@ -10,12 +10,12 @@ You need to prepare a Jeston development hardware. If you need TensorRT, you nee ...@@ -10,12 +10,12 @@ You need to prepare a Jeston development hardware. If you need TensorRT, you nee
1. Install PaddlePaddle in Jeston 1. Install PaddlePaddle in Jeston
paddlepaddle download [link](https://www.paddlepaddle.org.cn/inference/user_guides/download_lib.html#python) The PaddlePaddle download [link](https://www.paddlepaddle.org.cn/inference/user_guides/download_lib.html#python)
Please select the appropriate installation package for your Jetpack version, cuda version, and trt version. Please select the appropriate installation package for your Jetpack version, cuda version, and trt version. Here, we download paddlepaddle_gpu-2.3.0rc0-cp36-cp36m-linux_aarch64.whl.
Install paddlepaddle: Install PaddlePaddle:
```shell ```shell
pip3 install -U paddlepaddle_gpu-*-cp36-cp36m-linux_aarch64.whl pip3 install -U paddlepaddle_gpu-2.3.0rc0-cp36-cp36m-linux_aarch64.whl
``` ```
...@@ -36,7 +36,7 @@ pip3 install -r requirements.txt ...@@ -36,7 +36,7 @@ pip3 install -r requirements.txt
## 2. Perform prediction ## 2. Perform prediction
Obtain the PPOCR model from the [document](../../doc/doc_en/ppocr_introduction_en.md) model library. The following takes the PP-OCRv3 model as an example to introduce the use of the PPOCR model on jeston: Obtain the PPOCR model from the [document](https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/ppocr_introduction.md#6-%E6%A8%A1%E5%9E%8B%E5%BA%93) model library. The following takes the PP-OCRv3 model as an example to introduce the use of the PPOCR model on jeston:
Download and unzip the PP-OCRv3 models. Download and unzip the PP-OCRv3 models.
``` ```
......
...@@ -49,20 +49,6 @@ ...@@ -49,20 +49,6 @@
<a name="4-1"></a> <a name="4-1"></a>
### 4.1 Python推理 ### 4.1 Python推理
#### 4.1.1 基于训练模型的推理
训练完模型后可直接进行推理,以基于Resnet50_vd骨干网络,在ICDAR2015英文数据集训练的模型为例( 训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar)),使用方式如下:
```shell
# 下载模型并解压
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar
tar xf det_r50_vd_east_v2.0_train.tar
# 执行预测
python3 tools/infer_det.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.infer_img=./doc/imgs_en/img_10.jpg
```
#### 4.1.2 基于paddle预测引擎的推理
首先将EAST文本检测训练过程中保存的模型,转换成inference model。以基于Resnet50_vd骨干网络,在ICDAR2015英文数据集训练的模型为例( 训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar)),可以使用如下命令进行转换: 首先将EAST文本检测训练过程中保存的模型,转换成inference model。以基于Resnet50_vd骨干网络,在ICDAR2015英文数据集训练的模型为例( 训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar)),可以使用如下命令进行转换:
```shell ```shell
......
...@@ -38,7 +38,7 @@ PP-OCRv2在PP-OCR的基础上,进一步在5个方面重点优化,检测模 ...@@ -38,7 +38,7 @@ PP-OCRv2在PP-OCR的基础上,进一步在5个方面重点优化,检测模
#### PP-OCRv3 #### PP-OCRv3
PP-OCRv3在PP-OCRv2的基础上进一步升级。首先,文本检测采用改进fpn网络结构提升检测精度。 PP-OCRv3在PP-OCRv2的基础上进一步升级。文本检测采用改进fpn网络结构提升检测精度。
<a name="2"></a> <a name="2"></a>
......
...@@ -50,19 +50,6 @@ After the data download is complete, please refer to [Text Detection Training Tu ...@@ -50,19 +50,6 @@ After the data download is complete, please refer to [Text Detection Training Tu
<a name="4-1"></a> <a name="4-1"></a>
### 4.1 Python Inference ### 4.1 Python Inference
#### 4.1.2 Inference for Trained Model
After training, inference can be performed directly. Take [the model trained on the ICDAR2015 English dataset](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar) based on the Resnet50_vd backbone network as an example.
```shell
# download trained model
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar
tar xf det_r50_vd_east_v2.0_train.tar
# run inference
python3 tools/infer_det.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.infer_img=./doc/imgs_en/img_10.jpg
```
#### 4.1.2 Inference for Inference Model
First, convert the model saved in the EAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar)), you can use the following command to convert: First, convert the model saved in the EAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_east_v2.0_train.tar)), you can use the following command to convert:
```shell ```shell
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册