update docs
This commit is contained in:
licx
parent
6c11bce9e4
commit
ba58a2a631
|
|
@ -110,6 +110,9 @@ PaddleOCR开源的文本检测算法列表：
|
|||
|-|-|-|-|-|-|
|
||||
|SAST|ResNet50_vd|88.74%|79.80%|84.03%|[下载链接](https://paddleocr.bj.bcebos.com/SAST/sast_r50_vd_total_text.tar)|
|
||||
|
||||
**说明：** SAST模型训练额外加入了icdar2013、icdar2017、COCO-Text、ArT等公开数据集进行调优。PaddleOCR用到的经过整理格式的英文公开数据集下载：[百度云地址](https://pan.baidu.com/s/12cPnZcVuV1zn5DOd4mqjVw) (提取码: 2bpi)
|
||||
|
||||
|
||||
使用[LSVT](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_ch/datasets.md#1icdar2019-lsvt)街景数据集共3w张数据，训练中文检测模型的相关配置和预训练文件如下：
|
||||
|
||||
|模型|骨干网络|配置文件|预训练模型|
|
||||
|
|
@ -145,8 +148,7 @@ PaddleOCR开源的文本识别算法列表：
|
|||
|RARE|MobileNetV3|83.32%|rec_mv3_tps_bilstm_attn|[下载链接](https://paddleocr.bj.bcebos.com/rec_mv3_tps_bilstm_attn.tar)|
|
||||
|SRN|Resnet50_vd_fpn|88.33%|rec_r50fpn_vd_none_srn|[下载链接](https://paddleocr.bj.bcebos.com/SRN/rec_r50fpn_vd_none_srn.tar)|
|
||||
|
||||
**说明：** SRN模型使用了数据扰动方法对上述提到对两个训练集进行增广,增广后的数据可以在[百度网盘](todo)上下载。
|
||||
原始论文使用两阶段训练平均精度为89.74%，PaddleOCR中使用one-stage训练，平均精度为88.33%。两种预训练权重均在[下载链接](https://paddleocr.bj.bcebos.com/SRN/rec_r50fpn_vd_none_srn.tar)中。
|
||||
**说明：** SRN模型使用了数据扰动方法对上述提到对两个训练集进行增广，增广后的数据可以在[百度网盘](todo)上下载。原始论文使用两阶段训练平均精度为89.74%，PaddleOCR中使用one-stage训练，平均精度为88.33%。两种预训练权重均在[下载链接](https://paddleocr.bj.bcebos.com/SRN/rec_r50fpn_vd_none_srn.tar)中。
|
||||
|
||||
使用[LSVT](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_ch/datasets.md#1icdar2019-lsvt)街景数据集根据真值将图crop出来30w数据，进行位置校准。此外基于LSVT语料生成500w合成数据训练中文模型，相关配置和预训练文件如下：
|
||||
|
||||
|
|
|
|||
|
|
@ -1,13 +1,13 @@
|
|||
# 文字检测
|
||||
|
||||
本节以icdar15数据集为例，介绍PaddleOCR中检测模型的训练、评估与测试。
|
||||
本节以icdar2015数据集为例，介绍PaddleOCR中检测模型的训练、评估与测试。
|
||||
|
||||
## 数据准备
|
||||
icdar2015数据集可以从[官网](https://rrc.cvc.uab.es/?ch=4&com=downloads)下载到，首次下载需注册。
|
||||
|
||||
将下载到的数据集解压到工作目录下，假设解压在 PaddleOCR/train_data/ 下。另外，PaddleOCR将零散的标注文件整理成单独的标注文件
|
||||
，您可以通过wget的方式进行下载。
|
||||
```
|
||||
```shell
|
||||
# 在PaddleOCR路径下
|
||||
cd PaddleOCR/
|
||||
wget -P ./train_data/ https://paddleocr.bj.bcebos.com/dataset/train_icdar2015_label.txt
|
||||
|
|
@ -23,21 +23,21 @@ wget -P ./train_data/ https://paddleocr.bj.bcebos.com/dataset/test_icdar2015_la
|
|||
└─ test_icdar2015_label.txt icdar数据集的测试标注
|
||||
```
|
||||
|
||||
提供的标注文件格式为，其中中间是"\t"分隔：
|
||||
提供的标注文件格式如下，中间用"\t"分隔：
|
||||
```
|
||||
" 图像文件名 json.dumps编码的图像标注信息"
|
||||
ch4_test_images/img_61.jpg [{"transcription": "MASA", "points": [[310, 104], [416, 141], [418, 216], [312, 179]]}, {...}]
|
||||
```
|
||||
json.dumps编码前的图像标注信息是包含多个字典的list，字典中的 `points` 表示文本框的四个点的坐标(x, y)，从左上角的点开始顺时针排列。
|
||||
`transcription` 表示当前文本框的文字，在文本检测任务中并不需要这个信息。
|
||||
如果您想在其他数据集上训练PaddleOCR，可以按照上述形式构建标注文件。
|
||||
`transcription` 表示当前文本框的文字，**当其内容为“###”时，表示该文本框无效，在训练时会跳过。**
|
||||
|
||||
如果您想在其他数据集上训练，可以按照上述形式构建标注文件。
|
||||
|
||||
## 快速启动训练
|
||||
|
||||
首先下载模型backbone的pretrain model，PaddleOCR的检测模型目前支持两种backbone，分别是MobileNetV3、ResNet50_vd，
|
||||
您可以根据需求使用[PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/master/ppcls/modeling/architectures)中的模型更换backbone。
|
||||
```
|
||||
```shell
|
||||
cd PaddleOCR/
|
||||
# 下载MobileNetV3的预训练模型
|
||||
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x0_5_pretrained.tar
|
||||
|
|
@ -45,7 +45,7 @@ wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/Mob
|
|||
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_pretrained.tar
|
||||
|
||||
# 解压预训练模型文件，以MobileNetV3为例
|
||||
tar xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models/
|
||||
tar -xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models/
|
||||
|
||||
# 注：正确解压backbone预训练权重文件后，文件夹下包含众多以网络层命名的权重文件，格式如下：
|
||||
./pretrain_models/MobileNetV3_large_x0_5_pretrained/
|
||||
|
|
@ -57,11 +57,11 @@ tar xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models
|
|||
|
||||
```
|
||||
|
||||
**启动训练**
|
||||
#### 启动训练
|
||||
|
||||
*如果您安装的是cpu版本，请将配置文件中的 `use_gpu` 字段修改为false*
|
||||
|
||||
```
|
||||
```python
|
||||
python3 tools/train.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights=./pretrain_models/MobileNetV3_large_x0_5_pretrained/
|
||||
```
|
||||
|
||||
|
|
@ -69,52 +69,52 @@ python3 tools/train.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights=
|
|||
有关配置文件的详细解释，请参考[链接](./config.md)。
|
||||
|
||||
您也可以通过-o参数在不需要修改yml文件的情况下，改变训练的参数，比如，调整训练的学习率为0.0001
|
||||
```
|
||||
```python
|
||||
python3 tools/train.py -c configs/det/det_mv3_db.yml -o Optimizer.base_lr=0.0001
|
||||
```
|
||||
|
||||
**断点训练**
|
||||
#### 断点训练
|
||||
|
||||
如果训练程序中断，如果希望加载训练中断的模型从而恢复训练，可以通过指定Global.checkpoints指定要加载的模型路径：
|
||||
```
|
||||
```python
|
||||
python3 tools/train.py -c configs/det/det_mv3_db.yml -o Global.checkpoints=./your/trained/model
|
||||
```
|
||||
|
||||
**注意**：Global.checkpoints的优先级高于Global.pretrain_weights的优先级，即同时指定两个参数时，优先加载Global.checkpoints指定的模型，如果Global.checkpoints指定的模型路径有误，会加载Global.pretrain_weights指定的模型。
|
||||
**注意**：`Global.checkpoints`的优先级高于`Global.pretrain_weights`的优先级，即同时指定两个参数时，优先加载`Global.checkpoints`指定的模型，如果`Global.checkpoints`指定的模型路径有误，会加载`Global.pretrain_weights`指定的模型。
|
||||
|
||||
## 指标评估
|
||||
|
||||
PaddleOCR计算三个OCR检测相关的指标，分别是：Precision、Recall、Hmean。
|
||||
|
||||
运行如下代码，根据配置文件det_db_mv3.yml中save_res_path指定的测试集检测结果文件，计算评估指标。
|
||||
运行如下代码，根据配置文件`det_db_mv3.yml`中`save_res_path`指定的测试集检测结果文件，计算评估指标。
|
||||
|
||||
评估时设置后处理参数box_thresh=0.6，unclip_ratio=1.5，使用不同数据集、不同模型训练，可调整这两个参数进行优化
|
||||
```
|
||||
评估时设置后处理参数`box_thresh=0.6`，`unclip_ratio=1.5`，使用不同数据集、不同模型训练，可调整这两个参数进行优化
|
||||
```python
|
||||
python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{path/to/weights}/best_accuracy" PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
|
||||
```
|
||||
训练中模型参数默认保存在Global.save_model_dir目录下。在评估指标时，需要设置Global.checkpoints指向保存的参数文件。
|
||||
训练中模型参数默认保存在`Global.save_model_dir`目录下。在评估指标时，需要设置`Global.checkpoints`指向保存的参数文件。
|
||||
|
||||
比如：
|
||||
```
|
||||
```python
|
||||
python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="./output/det_db/best_accuracy" PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
|
||||
```
|
||||
|
||||
* 注：box_thresh、unclip_ratio是DB后处理所需要的参数，在评估EAST模型时不需要设置
|
||||
* 注：`box_thresh`、`unclip_ratio`是DB后处理所需要的参数，在评估EAST模型时不需要设置
|
||||
|
||||
## 测试检测效果
|
||||
|
||||
测试单张图像的检测效果
|
||||
```
|
||||
```python
|
||||
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o TestReader.infer_img="./doc/imgs_en/img_10.jpg" Global.checkpoints="./output/det_db/best_accuracy"
|
||||
```
|
||||
|
||||
测试DB模型时，调整后处理阈值，
|
||||
```
|
||||
```python
|
||||
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o TestReader.infer_img="./doc/imgs_en/img_10.jpg" Global.checkpoints="./output/det_db/best_accuracy" PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
|
||||
```
|
||||
|
||||
|
||||
测试文件夹下所有图像的检测效果
|
||||
```
|
||||
```python
|
||||
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o TestReader.infer_img="./doc/imgs_en/" Global.checkpoints="./output/det_db/best_accuracy"
|
||||
```
|
||||
|
|
|
|||
|
|
@ -1,6 +1,6 @@
|
|||
# TEXT DETECTION
|
||||
|
||||
This section uses the icdar15 dataset as an example to introduce the training, evaluation, and testing of the detection model in PaddleOCR.
|
||||
This section uses the icdar2015 dataset as an example to introduce the training, evaluation, and testing of the detection model in PaddleOCR.
|
||||
|
||||
## DATA PREPARATION
|
||||
The icdar2015 dataset can be obtained from [official website](https://rrc.cvc.uab.es/?ch=4&com=downloads). Registration is required for downloading.
|
||||
|
|
@ -27,10 +27,13 @@ The provided annotation file format is as follow, seperated by "\t":
|
|||
" Image file name Image annotation information encoded by json.dumps"
|
||||
ch4_test_images/img_61.jpg [{"transcription": "MASA", "points": [[310, 104], [416, 141], [418, 216], [312, 179]]}, {...}]
|
||||
```
|
||||
The image annotation after json.dumps() encoding is a list containing multiple dictionaries. The `points` in the dictionary represent the coordinates (x, y) of the four points of the text box, arranged clockwise from the point at the upper left corner.
|
||||
The image annotation after **json.dumps()** encoding is a list containing multiple dictionaries.
|
||||
|
||||
`transcription` represents the text of the current text box, and this information is not needed in the text detection task.
|
||||
If you want to train PaddleOCR on other datasets, you can build the annotation file according to the above format.
|
||||
The `points` in the dictionary represent the coordinates (x, y) of the four points of the text box, arranged clockwise from the point at the upper left corner.
|
||||
|
||||
`transcription` represents the text of the current text box. **When its content is "###" it means that the text box is invalid and will be skipped during training.**
|
||||
|
||||
If you want to train PaddleOCR on other datasets, please build the annotation file according to the above format.
|
||||
|
||||
|
||||
## TRAINING
|
||||
|
|
@ -56,7 +59,7 @@ tar xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models
|
|||
|
||||
```
|
||||
|
||||
**START TRAINING**
|
||||
#### START TRAINING
|
||||
*If CPU version installed, please set the parameter `use_gpu` to `false` in the configuration.*
|
||||
```
|
||||
python3 tools/train.py -c configs/det/det_mv3_db.yml
|
||||
|
|
@ -70,7 +73,7 @@ You can also use `-o` to change the training parameters without modifying the ym
|
|||
python3 tools/train.py -c configs/det/det_mv3_db.yml -o Optimizer.base_lr=0.0001
|
||||
```
|
||||
|
||||
**load trained model and conntinue training**
|
||||
#### load trained model and conntinue training
|
||||
If you expect to load trained model and continue the training again, you can specify the parameter `Global.checkpoints` as the model path to be loaded.
|
||||
|
||||
For example:
|
||||
|
|
|
|||
Loading…
Reference in New Issue