Unverified Commit 6893d151 authored by Thomas Young's avatar Thomas Young Committed by GitHub
Browse files

Merge pull request #1 from PaddlePaddle/dygraph

Dygraph
parents 32665fe5 58794e06
......@@ -104,27 +104,16 @@ python3 generate_multi_language_configs.py -l it \
| german_mobile_v2.0_rec |德文识别|[rec_german_lite_train.yml](../../configs/rec/multi_language/rec_german_lite_train.yml)|2.65M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/german_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/german_mobile_v2.0_rec_train.tar) |
| korean_mobile_v2.0_rec |韩文识别|[rec_korean_lite_train.yml](../../configs/rec/multi_language/rec_korean_lite_train.yml)|3.9M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/korean_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/korean_mobile_v2.0_rec_train.tar) |
| japan_mobile_v2.0_rec |日文识别|[rec_japan_lite_train.yml](../../configs/rec/multi_language/rec_japan_lite_train.yml)|4.23M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/japan_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/japan_mobile_v2.0_rec_train.tar) |
| it_mobile_v2.0_rec |意大利文识别|rec_it_lite_train.yml|2.53M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/it_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/it_mobile_v2.0_rec_train.tar) |
| xi_mobile_v2.0_rec |西班牙文识别|rec_xi_lite_train.yml|2.53M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/xi_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/xi_mobile_v2.0_rec_train.tar) |
| pu_mobile_v2.0_rec |葡萄牙文识别|rec_pu_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/pu_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/pu_mobile_v2.0_rec_train.tar) |
| ru_mobile_v2.0_rec |俄罗斯文识别|rec_ru_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ru_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ru_mobile_v2.0_rec_train.tar) |
| ar_mobile_v2.0_rec |阿拉伯文识别|rec_ar_lite_train.yml|2.53M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ar_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ar_mobile_v2.0_rec_train.tar) |
| hi_mobile_v2.0_rec |印地文识别|rec_hi_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/hi_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/hi_mobile_v2.0_rec_train.tar) |
| chinese_cht_mobile_v2.0_rec |中文繁体识别|rec_chinese_cht_lite_train.yml|5.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/chinese_cht_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/chinese_cht_mobile_v2.0_rec_train.tar) |
| ug_mobile_v2.0_rec |维吾尔文识别|rec_ug_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ug_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ug_mobile_v2.0_rec_train.tar) |
| fa_mobile_v2.0_rec |波斯文识别|rec_fa_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/fa_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/fa_mobile_v2.0_rec_train.tar) |
| ur_mobile_v2.0_rec |乌尔都文识别|rec_ur_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ur_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ur_mobile_v2.0_rec_train.tar) |
| rs_mobile_v2.0_rec |塞尔维亚文(latin)识别|rec_rs_lite_train.yml|2.53M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/rs_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/rs_mobile_v2.0_rec_train.tar) |
| oc_mobile_v2.0_rec |欧西坦文识别|rec_oc_lite_train.yml|2.53M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/oc_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/oc_mobile_v2.0_rec_train.tar) |
| mr_mobile_v2.0_rec |马拉地文识别|rec_mr_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/mr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/mr_mobile_v2.0_rec_train.tar) |
| ne_mobile_v2.0_rec |尼泊尔文识别|rec_ne_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ne_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ne_mobile_v2.0_rec_train.tar) |
| rsc_mobile_v2.0_rec |塞尔维亚文(cyrillic)识别|rec_rsc_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/rsc_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/rsc_mobile_v2.0_rec_train.tar) |
| bg_mobile_v2.0_rec |保加利亚文识别|rec_bg_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/bg_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/bg_mobile_v2.0_rec_train.tar) |
| uk_mobile_v2.0_rec |乌克兰文识别|rec_uk_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/uk_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/uk_mobile_v2.0_rec_train.tar) |
| be_mobile_v2.0_rec |白俄罗斯文识别|rec_be_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/be_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/be_mobile_v2.0_rec_train.tar) |
| te_mobile_v2.0_rec |泰卢固文识别|rec_te_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/te_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/te_mobile_v2.0_rec_train.tar) |
| ka_mobile_v2.0_rec |卡纳达文识别|rec_ka_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ka_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ka_mobile_v2.0_rec_train.tar) |
| ta_mobile_v2.0_rec |泰米尔文识别|rec_ta_lite_train.yml|2.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ta_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ta_mobile_v2.0_rec_train.tar) |
| latin_mobile_v2.0_rec | 拉丁文识别 | [rec_latin_lite_train.yml](../../configs/rec/multi_language/rec_latin_lite_train.yml) |2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/latin_ppocr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/latin_ppocr_mobile_v2.0_rec_train.tar) |
| arabic_mobile_v2.0_rec | 阿拉伯字母 | [rec_arabic_lite_train.yml](../../configs/rec/multi_language/rec_arabic_lite_train.yml) |2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/arabic_ppocr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/arabic_ppocr_mobile_v2.0_rec_train.tar) |
| cyrillic_mobile_v2.0_rec | 斯拉夫字母 | [rec_cyrillic_lite_train.yml](../../configs/rec/multi_language/rec_cyrillic_lite_train.yml) |2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/cyrillic_ppocr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/cyrillic_ppocr_mobile_v2.0_rec_train.tar) |
| devanagari_mobile_v2.0_rec | 梵文字母 | [rec_devanagari_lite_train.yml](../../configs/rec/multi_language/rec_devanagari_lite_train.yml) |2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/devanagari_ppocr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/devanagari_ppocr_mobile_v2.0_rec_train.tar) |
更多支持语种请参考: [多语言模型](./multi_languages.md)
<a name="文本方向分类模型"></a>
......
# 多语言模型
**近期更新**
- 2021.4.9 支持**80种**语言的检测和识别
- 2021.4.9 支持**轻量高精度**英文模型检测识别
PaddleOCR 旨在打造一套丰富、领先、且实用的OCR工具库,不仅提供了通用场景下的中英文模型,也提供了专门在英文场景下训练的模型,
和覆盖[80个语言](#语种缩写)的小语种模型。
其中英文模型支持,大小写字母和常见标点的检测识别,并优化了空格字符的识别:
<div align="center">
<img src="../imgs_results/multi_lang/img_12.jpg" width="900" height="300">
</div>
小语种模型覆盖了拉丁语系、阿拉伯语系、中文繁体、韩语、日语等等:
<div align="center">
<img src="../imgs_results/multi_lang/japan_2.jpg" width="600" height="300">
<img src="../imgs_results/multi_lang/french_0.jpg" width="300" height="300">
<img src="../imgs_results/multi_lang/korean_0.jpg" width="500" height="300">
<img src="../imgs_results/multi_lang/arabic_0.jpg" width="300" height="300">
</div>
本文档将简要介绍小语种模型的使用方法。
- [1 安装](#安装)
- [1.1 paddle 安装](#paddle安装)
- [1.2 paddleocr package 安装](#paddleocr_package_安装)
- [2 快速使用](#快速使用)
- [2.1 命令行运行](#命令行运行)
- [2.2 python 脚本运行](#python_脚本运行)
- [3 自定义训练](#自定义训练)
- [4 预测部署](#预测部署)
- [4 支持语种及缩写](#语种缩写)
<a name="安装"></a>
## 1 安装
<a name="paddle安装"></a>
### 1.1 paddle 安装
```
# cpu
pip install paddlepaddle
# gpu
pip install paddlepaddle-gpu
```
<a name="paddleocr_package_安装"></a>
### 1.2 paddleocr package 安装
pip 安装
```
pip install "paddleocr>=2.0.6" # 推荐使用2.0.6版本
```
本地构建并安装
```
python3 setup.py bdist_wheel
pip3 install dist/paddleocr-x.x.x-py3-none-any.whl # x.x.x是paddleocr的版本号
```
<a name="快速使用"></a>
## 2 快速使用
<a name="命令行运行"></a>
### 2.1 命令行运行
查看帮助信息
```
paddleocr -h
```
* 整图预测(检测+识别)
Paddleocr目前支持80个语种,可以通过修改--lang参数进行切换,具体支持的[语种](#语种缩写)可查看表格。
``` bash
paddleocr --image_dir doc/imgs_en/254.jpg --lang=en
```
<div align="center">
<img src="../imgs_en/254.jpg" width="300" height="600">
<img src="../imgs_results/multi_lang/img_02.jpg" width="600" height="600">
</div>
结果是一个list,每个item包含了文本框,文字和识别置信度
```text
[('PHO CAPITAL', 0.95723116), [[66.0, 50.0], [327.0, 44.0], [327.0, 76.0], [67.0, 82.0]]]
[('107 State Street', 0.96311164), [[72.0, 90.0], [451.0, 84.0], [452.0, 116.0], [73.0, 121.0]]]
[('Montpelier Vermont', 0.97389287), [[69.0, 132.0], [501.0, 126.0], [501.0, 158.0], [70.0, 164.0]]]
[('8022256183', 0.99810505), [[71.0, 175.0], [363.0, 170.0], [364.0, 202.0], [72.0, 207.0]]]
[('REG 07-24-201706:59 PM', 0.93537045), [[73.0, 299.0], [653.0, 281.0], [654.0, 318.0], [74.0, 336.0]]]
[('045555', 0.99346405), [[509.0, 331.0], [651.0, 325.0], [652.0, 356.0], [511.0, 362.0]]]
[('CT1', 0.9988654), [[535.0, 367.0], [654.0, 367.0], [654.0, 406.0], [535.0, 406.0]]]
......
```
* 识别预测
```bash
paddleocr --image_dir doc/imgs_words_en/word_308.png --det false --lang=en
```
结果是一个tuple,返回识别结果和识别置信度
```text
(0.99879867, 'LITTLE')
```
* 检测预测
```
paddleocr --image_dir PaddleOCR/doc/imgs/11.jpg --rec false
```
结果是一个list,每个item只包含文本框
```
[[26.0, 457.0], [137.0, 457.0], [137.0, 477.0], [26.0, 477.0]]
[[25.0, 425.0], [372.0, 425.0], [372.0, 448.0], [25.0, 448.0]]
[[128.0, 397.0], [273.0, 397.0], [273.0, 414.0], [128.0, 414.0]]
......
```
<a name="python_脚本运行"></a>
### 2.2 python 脚本运行
ppocr 也支持在python脚本中运行,便于嵌入到您自己的代码中 :
* 整图预测(检测+识别)
```
from paddleocr import PaddleOCR, draw_ocr
# 同样也是通过修改 lang 参数切换语种
ocr = PaddleOCR(lang="korean") # 首次执行会自动下载模型文件
img_path = 'doc/imgs/korean_1.jpg '
result = ocr.ocr(img_path)
# 可通过参数控制单独执行识别、检测
# result = ocr.ocr(img_path, det=False) 只执行识别
# result = ocr.ocr(img_path, rec=False) 只执行检测
# 打印检测框和识别结果
for line in result:
print(line)
# 可视化
from PIL import Image
image = Image.open(img_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores, font_path='/path/to/PaddleOCR/doc/fonts/korean.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')
```
结果可视化:
<div align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/release/2.1/doc/imgs_results/korean.jpg" width="800">
</div>
ppocr 还支持方向分类, 更多使用方式请参考:[whl包使用说明](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.0/doc/doc_ch/whl.md)
<a name="自定义训练"></a>
## 3 自定义训练
ppocr 支持使用自己的数据进行自定义训练或finetune, 其中识别模型可以参考 [法语配置文件](../../configs/rec/multi_language/rec_french_lite_train.yml)
修改训练数据路径、字典等参数。
具体数据准备、训练过程可参考:[文本检测](../doc_ch/detection.md)[文本识别](../doc_ch/recognition.md),更多功能如预测部署、
数据标注等功能可以阅读完整的[文档教程](../../README_ch.md)
<a name="预测部署"></a>
## 4 预测部署
除了安装whl包进行快速预测,ppocr 也提供了多种预测部署方式,如有需求可阅读相关文档:
- [基于Python脚本预测引擎推理](./inference.md)
- [基于C++预测引擎推理](../../deploy/cpp_infer/readme.md)
- [服务化部署](../../deploy/hubserving/readme.md)
- [端侧部署](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/lite/readme.md)
- [Benchmark](./benchmark.md)
<a name="语种缩写"></a>
## 5 支持语种及缩写
| 语种 | 描述 | 缩写 | | 语种 | 描述 | 缩写 |
| --- | --- | --- | ---|--- | --- | --- |
|中文|chinese and english|ch| |保加利亚文|Bulgarian |bg|
|英文|english|en| |乌克兰文|Ukranian|uk|
|法文|french|fr| |白俄罗斯文|Belarusian|be|
|德文|german|german| |泰卢固文|Telugu |te|
|日文|japan|japan| | |阿巴扎文|Abaza |abq|
|韩文|korean|korean| |泰米尔文|Tamil |ta|
|中文繁体|chinese traditional |ch_tra| |南非荷兰文 |Afrikaans |af|
|意大利文| Italian |it| |阿塞拜疆文 |Azerbaijani |az|
|西班牙文|Spanish |es| |波斯尼亚文|Bosnian|bs|
|葡萄牙文| Portuguese|pt| |捷克文|Czech|cs|
|俄罗斯文|Russia|ru| |威尔士文 |Welsh |cy|
|阿拉伯文|Arabic|ar| |丹麦文 |Danish|da|
|印地文|Hindi|hi| |爱沙尼亚文 |Estonian |et|
|维吾尔|Uyghur|ug| |爱尔兰文 |Irish |ga|
|波斯文|Persian|fa| |克罗地亚文|Croatian |hr|
|乌尔都文|Urdu|ur| |匈牙利文|Hungarian |hu|
|塞尔维亚文(latin)| Serbian(latin) |rs_latin| |印尼文|Indonesian|id|
|欧西坦文|Occitan |oc| |冰岛文 |Icelandic|is|
|马拉地文|Marathi|mr| |库尔德文 |Kurdish|ku|
|尼泊尔文|Nepali|ne| |立陶宛文|Lithuanian |lt|
|塞尔维亚文(cyrillic)|Serbian(cyrillic)|rs_cyrillic| |拉脱维亚文 |Latvian |lv|
|毛利文|Maori|mi| | 达尔瓦文|Dargwa |dar|
|马来文 |Malay|ms| | 因古什文|Ingush |inh|
|马耳他文 |Maltese |mt| | 拉克文|Lak |lbe|
|荷兰文 |Dutch |nl| | 莱兹甘文|Lezghian |lez|
|挪威文 |Norwegian |no| |塔巴萨兰文 |Tabassaran |tab|
|波兰文|Polish |pl| | 比尔哈文|Bihari |bh|
| 罗马尼亚文|Romanian |ro| | 迈蒂利文|Maithili |mai|
| 斯洛伐克文|Slovak |sk| | 昂加文|Angika |ang|
| 斯洛文尼亚文|Slovenian |sl| | 孟加拉文|Bhojpuri |bho|
| 阿尔巴尼亚文|Albanian |sq| | 摩揭陀文 |Magahi |mah|
| 瑞典文|Swedish |sv| | 那格浦尔文|Nagpur |sck|
| 西瓦希里文|Swahili |sw| | 尼瓦尔文|Newari |new|
| 塔加洛文|Tagalog |tl| | 保加利亚文 |Goan Konkani|gom|
| 土耳其文|Turkish |tr| | 沙特阿拉伯文|Saudi Arabia|sa|
| 乌兹别克文|Uzbek |uz| | 阿瓦尔文|Avar |ava|
| 越南文|Vietnamese |vi| | 阿瓦尔文|Avar |ava|
| 蒙古文|Mongolian |mn| | 阿迪赫文|Adyghe |ady|
# 端对端OCR算法-PGNet
- [一、简介](#简介)
- [二、环境配置](#环境配置)
- [三、快速使用](#快速使用)
- [四、模型训练、评估、推理](#模型训练、评估、推理)
<a name="简介"></a>
## 一、简介
OCR算法可以分为两阶段算法和端对端的算法。二阶段OCR算法一般分为两个部分,文本检测和文本识别算法,文件检测算法从图像中得到文本行的检测框,然后识别算法去识别文本框中的内容。而端对端OCR算法可以在一个算法中完成文字检测和文字识别,其基本思想是设计一个同时具有检测单元和识别模块的模型,共享其中两者的CNN特征,并联合训练。由于一个算法即可完成文字识别,端对端模型更小,速度更快。
### PGNet算法介绍
近些年来,端对端OCR算法得到了良好的发展,包括MaskTextSpotter系列、TextSnake、TextDragon、PGNet系列等算法。在这些算法中,PGNet算法具备其他算法不具备的优势,包括:
- 设计PGNet loss指导训练,不需要字符级别的标注
- 不需要NMS和ROI相关操作,加速预测
- 提出预测文本行内的阅读顺序模块;
- 提出基于图的修正模块(GRM)来进一步提高模型识别性能
- 精度更高,预测速度更快
PGNet算法细节详见[论文](https://www.aaai.org/AAAI21Papers/AAAI-2885.WangP.pdf) ,算法原理图如下所示:
![](../pgnet_framework.png)
输入图像经过特征提取送入四个分支,分别是:文本边缘偏移量预测TBO模块,文本中心线预测TCL模块,文本方向偏移量预测TDO模块,以及文本字符分类图预测TCC模块。
其中TBO以及TCL的输出经过后处理后可以得到文本的检测结果,TCL、TDO、TCC负责文本识别。
其检测识别效果图如下:
![](../imgs_results/e2e_res_img293_pgnet.png)
![](../imgs_results/e2e_res_img295_pgnet.png)
### 性能指标
测试集: Total Text
测试环境: NVIDIA Tesla V100-SXM2-16GB
|PGNetA|det_precision|det_recall|det_f_score|e2e_precision|e2e_recall|e2e_f_score|FPS|下载|
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
|Paper|85.30|86.80|86.1|-|-|61.7|38.20 (size=640)|-|
|Ours|87.03|82.48|84.69|61.71|58.43|60.03|48.73 (size=768)|[下载链接](https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/en_server_pgnetA.tar)|
*note:PaddleOCR里的PGNet实现针对预测速度做了优化,在精度下降可接受范围内,可以显著提升端对端预测速度*
<a name="环境配置"></a>
## 二、环境配置
请先参考[快速安装](./installation.md)配置PaddleOCR运行环境。
<a name="快速使用"></a>
## 三、快速使用
### inference模型下载
本节以训练好的端到端模型为例,快速使用模型预测,首先下载训练好的端到端inference模型[下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/e2e_server_pgnetA_infer.tar)
```
mkdir inference && cd inference
# 下载英文端到端模型并解压
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/e2e_server_pgnetA_infer.tar && tar xf e2e_server_pgnetA_infer.tar
```
* windows 环境下如果没有安装wget,下载模型时可将链接复制到浏览器中下载,并解压放置在相应目录下
解压完毕后应有如下文件结构:
```
├── e2e_server_pgnetA_infer
│ ├── inference.pdiparams
│ ├── inference.pdiparams.info
│ └── inference.pdmodel
```
### 单张图像或者图像集合预测
```bash
# 预测image_dir指定的单张图像
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/img623.jpg" --e2e_model_dir="./inference/e2e_server_pgnetA_infer/" --e2e_pgnet_polygon=True
# 预测image_dir指定的图像集合
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/" --e2e_model_dir="./inference/e2e_server_pgnetA_infer/" --e2e_pgnet_polygon=True
# 如果想使用CPU进行预测,需设置use_gpu参数为False
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/img623.jpg" --e2e_model_dir="./inference/e2e_server_pgnetA_infer/" --e2e_pgnet_polygon=True --use_gpu=False
```
### 可视化结果
可视化文本检测结果默认保存到./inference_results文件夹里面,结果文件的名称前缀为'e2e_res'。结果示例如下:
![](../imgs_results/e2e_res_img623_pgnet.jpg)
<a name="模型训练、评估、推理"></a>
## 四、模型训练、评估、推理
本节以totaltext数据集为例,介绍PaddleOCR中端到端模型的训练、评估与测试。
### 准备数据
下载解压[totaltext](https://paddleocr.bj.bcebos.com/dataset/total_text.tar) 数据集到PaddleOCR/train_data/目录,数据集组织结构:
```
/PaddleOCR/train_data/total_text/train/
|- rgb/ # total_text数据集的训练数据
|- img11.jpg
| ...
|- train.txt # total_text数据集的训练标注
```
total_text.txt标注文件格式如下,文件名和标注信息中间用"\t"分隔:
```
" 图像文件名 json.dumps编码的图像标注信息"
rgb/img11.jpg [{"transcription": "ASRAMA", "points": [[214.0, 325.0], [235.0, 308.0], [259.0, 296.0], [286.0, 291.0], [313.0, 295.0], [338.0, 305.0], [362.0, 320.0], [349.0, 347.0], [330.0, 337.0], [310.0, 329.0], [290.0, 324.0], [269.0, 328.0], [249.0, 336.0], [231.0, 346.0]]}, {...}]
```
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,**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。
### 启动训练
PGNet训练分为两个步骤:step1: 在合成数据上训练,得到预训练模型,此时模型精度依然较低;step2: 加载预训练模型,在totaltext数据集上训练;为快速训练,我们直接提供了step1的预训练模型。
```shell
cd PaddleOCR/
下载step1 预训练模型
wget -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/train_step1.tar
可以得到以下的文件格式
./pretrain_models/train_step1/
└─ best_accuracy.pdopt
└─ best_accuracy.states
└─ best_accuracy.pdparams
```
*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*
```shell
# 单机单卡训练 e2e 模型
python3 tools/train.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.pretrained_model=./pretrain_models/train_step1/best_accuracy Global.load_static_weights=False
# 单机多卡训练,通过 --gpus 参数设置使用的GPU ID
python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.pretrained_model=./pretrain_models/train_step1/best_accuracy Global.load_static_weights=False
```
上述指令中,通过-c 选择训练使用configs/e2e/e2e_r50_vd_pg.yml配置文件。
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
```shell
python3 tools/train.py -c configs/e2e/e2e_r50_vd_pg.yml -o Optimizer.base_lr=0.0001
```
#### 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:
```shell
python3 tools/train.py -c configs/e2e/e2e_r50_vd_pg.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。
PaddleOCR计算三个OCR端到端相关的指标,分别是:Precision、Recall、Hmean。
运行如下代码,根据配置文件`e2e_r50_vd_pg.yml``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`max_side_len=768`,使用不同数据集、不同模型训练,可调整参数进行优化
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。
```shell
python3 tools/eval.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.checkpoints="{path/to/weights}/best_accuracy"
```
### 模型预测
测试单张图像的端到端识别效果
```shell
python3 tools/infer_e2e.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/e2e_pgnet/best_accuracy" Global.load_static_weights=false
```
测试文件夹下所有图像的端到端识别效果
```shell
python3 tools/infer_e2e.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/e2e_pgnet/best_accuracy" Global.load_static_weights=false
```
### 预测推理
#### (1). 四边形文本检测模型(ICDAR2015)
首先将PGNet端到端训练过程中保存的模型,转换成inference model。以基于Resnet50_vd骨干网络,以英文数据集训练的模型为例[模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/en_server_pgnetA.tar) ,可以使用如下命令进行转换:
```
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/en_server_pgnetA.tar && tar xf en_server_pgnetA.tar
python3 tools/export_model.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.pretrained_model=./en_server_pgnetA/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/e2e
```
**PGNet端到端模型推理,需要设置参数`--e2e_algorithm="PGNet"`**,可以执行如下命令:
```
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/img_10.jpg" --e2e_model_dir="./inference/e2e/" --e2e_pgnet_polygon=False
```
可视化文本检测结果默认保存到`./inference_results`文件夹里面,结果文件的名称前缀为'e2e_res'。结果示例如下:
![](../imgs_results/e2e_res_img_10_pgnet.jpg)
#### (2). 弯曲文本检测模型(Total-Text)
对于弯曲文本样例
**PGNet端到端模型推理,需要设置参数`--e2e_algorithm="PGNet"`,同时,还需要增加参数`--e2e_pgnet_polygon=True`,**可以执行如下命令:
```
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/img623.jpg" --e2e_model_dir="./inference/e2e/" --e2e_pgnet_polygon=True
```
可视化文本端到端结果默认保存到`./inference_results`文件夹里面,结果文件的名称前缀为'e2e_res'。结果示例如下:
![](../imgs_results/e2e_res_img623_pgnet.jpg)
......@@ -138,7 +138,7 @@ PaddleOCR内置了一部分字典,可以按需使用。
`ppocr/utils/dict/german_dict.txt` 是一个包含131个字符的德文字典
`ppocr/utils/dict/en_dict.txt` 是一个包含63个字符的英文字典
`ppocr/utils/en_dict.txt` 是一个包含96个字符的英文字典
......@@ -285,7 +285,7 @@ Eval:
<a name="小语种"></a>
#### 2.3 小语种
PaddleOCR目前已支持26种(除中文外)语种识别,`configs/rec/multi_languages` 路径下提供了一个多语言的配置文件模版: [rec_multi_language_lite_train.yml](../../configs/rec/multi_language/rec_multi_language_lite_train.yml)
PaddleOCR目前已支持80种(除中文外)语种识别,`configs/rec/multi_languages` 路径下提供了一个多语言的配置文件模版: [rec_multi_language_lite_train.yml](../../configs/rec/multi_language/rec_multi_language_lite_train.yml)
您有两种方式创建所需的配置文件:
......@@ -368,26 +368,12 @@ PaddleOCR目前已支持26种(除中文外)语种识别,`configs/rec/multi
| rec_ger_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 德语 | german |
| rec_japan_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 日语 | japan |
| rec_korean_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 韩语 | korean |
| rec_it_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 意大利语 | it |
| rec_xi_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 西班牙语 | xi |
| rec_pu_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 葡萄牙语 | pu |
| rec_ru_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 俄罗斯语 | ru |
| rec_ar_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 阿拉伯语 | ar |
| rec_hi_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 印地语 | hi |
| rec_ug_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 维吾尔语 | ug |
| rec_fa_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 波斯语 | fa |
| rec_ur_ite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 乌尔都语 | ur |
| rec_rs_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 塞尔维亚(latin)语 | rs |
| rec_oc_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 欧西坦语 | oc |
| rec_mr_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 马拉地语 | mr |
| rec_ne_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 尼泊尔语 | ne |
| rec_rsc_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 塞尔维亚(cyrillic)语 | rsc |
| rec_bg_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 保加利亚语 | bg |
| rec_uk_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 乌克兰语 | uk |
| rec_be_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 白俄罗斯语 | be |
| rec_te_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 泰卢固语 | te |
| rec_ka_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 卡纳达语 | ka |
| rec_ta_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 泰米尔语 | ta |
| rec_latin_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 拉丁字母 | latin |
| rec_arabic_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 阿拉伯字母 | ar |
| rec_cyrillic_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 斯拉夫字母 | cyrillic |
| rec_devanagari_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | 梵文字母 | devanagari |
更多支持语种请参考: [多语言模型](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_ch/multi_languages.md#%E8%AF%AD%E7%A7%8D%E7%BC%A9%E5%86%99)
多语言模型训练方式与中文模型一致,训练数据集均为100w的合成数据,少量的字体可以在 [百度网盘](https://pan.baidu.com/s/1bS_u207Rm7YbY33wOECKDA) 上下载,提取码:frgi。
......
......@@ -38,28 +38,17 @@ If you want to train PaddleOCR on other datasets, please build the annotation fi
## TRAINING
First download the pretrained model. The detection model of PaddleOCR currently supports 3 backbones, namely MobileNetV3, ResNet18_vd and ResNet50_vd. You can use the model in [PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/master/ppcls/modeling/architectures) to replace backbone according to your needs.
First download the pretrained model. The detection model of PaddleOCR currently supports 3 backbones, namely MobileNetV3, ResNet18_vd and ResNet50_vd. You can use the model in [PaddleClas](https://github.com/PaddlePaddle/PaddleClas/tree/develop/ppcls/modeling/architectures) to replace backbone according to your needs.
And the responding download link of backbone pretrain weights can be found in [PaddleClas repo](https://github.com/PaddlePaddle/PaddleClas#mobile-series).
```shell
cd PaddleOCR/
# Download the pre-trained model of MobileNetV3
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/MobileNetV3_large_x0_5_pretrained.tar
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/MobileNetV3_large_x0_5_pretrained.pdparams
# or, download the pre-trained model of ResNet18_vd
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet18_vd_pretrained.tar
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet18_vd_pretrained.pdparams
# or, download the pre-trained model of ResNet50_vd
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_ssld_pretrained.tar
wget -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/ResNet50_vd_ssld_pretrained.pdparams
# decompressing the pre-training model file, take MobileNetV3 as an example
tar -xf ./pretrain_models/MobileNetV3_large_x0_5_pretrained.tar ./pretrain_models/
# Note: After decompressing the backbone pre-training weight file correctly, the file list in the folder is as follows:
./pretrain_models/MobileNetV3_large_x0_5_pretrained/
└─ conv_last_bn_mean
└─ conv_last_bn_offset
└─ conv_last_bn_scale
└─ conv_last_bn_variance
└─ ......
```
#### START TRAINING
*If CPU version installed, please set the parameter `use_gpu` to `false` in the configuration.*
......@@ -113,16 +102,16 @@ python3 tools/eval.py -c configs/det/det_mv3_db.yml -o Global.checkpoints="{pat
Test the detection result on a single image:
```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy"
```
When testing the DB model, adjust the post-processing threshold:
```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/det_db/best_accuracy" PostProcess.box_thresh=0.6 PostProcess.unclip_ratio=1.5
```
Test the detection result on all images in the folder:
```shell
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy" Global.load_static_weights=false
python3 tools/infer_det.py -c configs/det/det_mv3_db.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/det_db/best_accuracy"
```
......@@ -52,10 +52,9 @@ The above model is a DB algorithm trained with MobileNetV3 as the backbone. To c
# -c Set the training algorithm yml configuration file
# -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.load_static_weights needs to be set to False
# Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_db/
python3 tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_inference_dir=./inference/det_db/
```
When converting to an inference model, the configuration file used is the same as the configuration file used during training. In addition, you also need to set the `Global.pretrained_model` parameter in the configuration file.
......@@ -80,10 +79,9 @@ The recognition model is converted to the inference model in the same way as the
# -c Set the training algorithm yml configuration file
# -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.load_static_weights needs to be set to False
# Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_rec_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/rec_crnn/
python3 tools/export_model.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_rec_train/best_accuracy Global.save_inference_dir=./inference/rec_crnn/
```
If you have a model trained on your own dataset with a different dictionary file, please make sure that you modify the `character_dict_path` in the configuration file to your dictionary file path.
......@@ -109,10 +107,9 @@ The angle classification model is converted to the inference model in the same w
# -c Set the training algorithm yml configuration file
# -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.load_static_weights needs to be set to False
# Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_cls_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/cls/
python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_cls_train/best_accuracy Global.save_inference_dir=./inference/cls/
```
After the conversion is successful, there are two files in the directory:
......@@ -171,7 +168,7 @@ python3 tools/infer/predict_det.py --image_dir="./doc/imgs/22.jpg" --det_model_d
First, convert the model saved in the DB 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 an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_db_v2.0_train.tar)), you can use the following command to convert:
```
python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o Global.pretrained_model=./det_r50_vd_db_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_db
python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o Global.pretrained_model=./det_r50_vd_db_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_db
```
DB text detection model inference, you can execute the following command:
......@@ -192,7 +189,7 @@ The visualized text detection results are saved to the `./inference_results` fol
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 an 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:
```
python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_east
python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_east
```
**For EAST text detection model inference, you need to set the parameter ``--det_algorithm="EAST"``**, run the following command:
......@@ -213,7 +210,7 @@ The visualized text detection results are saved to the `./inference_results` fol
First, convert the model saved in the SAST 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 an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_icdar15_v2.0_train.tar)), you can use the following command to convert:
```
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o Global.pretrained_model=./det_r50_vd_sast_icdar15_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_sast_ic15
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o Global.pretrained_model=./det_r50_vd_sast_icdar15_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_sast_ic15
```
**For SAST quadrangle text detection model inference, you need to set the parameter `--det_algorithm="SAST"`**, run the following command:
......@@ -230,7 +227,7 @@ The visualized text detection results are saved to the `./inference_results` fol
First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the Total-Text English dataset as an example ([model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_sast_totaltext_v2.0_train.tar)), you can use the following command to convert:
```
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o Global.pretrained_model=./det_r50_vd_sast_totaltext_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/det_sast_tt
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o Global.pretrained_model=./det_r50_vd_sast_totaltext_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_sast_tt
```
**For SAST curved text detection model inference, you need to set the parameter `--det_algorithm="SAST"` and `--det_sast_polygon=True`**, run the following command:
......@@ -279,7 +276,7 @@ Taking CRNN as an example, we introduce the recognition model inference based on
First, convert the model saved in the CRNN text recognition training process into an inference model. Taking the model based on Resnet34_vd backbone network, using MJSynth and SynthText (two English text recognition synthetic datasets) for training, as an example ([model download address](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/rec_r34_vd_none_bilstm_ctc_v2.0_train.tar)). It can be converted as follow:
```
python3 tools/export_model.py -c configs/det/rec_r34_vd_none_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_none_bilstm_ctc_v2.0_train/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/rec_crnn
python3 tools/export_model.py -c configs/det/rec_r34_vd_none_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_none_bilstm_ctc_v2.0_train/best_accuracy Global.save_inference_dir=./inference/rec_crnn
```
For CRNN text recognition model inference, execute the following commands:
......@@ -379,14 +376,18 @@ After executing the command, the prediction results (classification angle and sc
<a name="LIGHTWEIGHT_CHINESE_MODEL"></a>
### 1. LIGHTWEIGHT CHINESE MODEL
When performing prediction, you need to specify the path of a single image or a folder of images through the parameter `image_dir`, the parameter `det_model_dir` specifies the path to detect the inference model, the parameter `cls_model_dir` specifies the path to angle classification inference model and the parameter `rec_model_dir` specifies the path to identify the inference model. The parameter `use_angle_cls` is used to control whether to enable the angle classification model.The visualized recognition results are saved to the `./inference_results` folder by default.
When performing prediction, you need to specify the path of a single image or a folder of images through the parameter `image_dir`, the parameter `det_model_dir` specifies the path to detect the inference model, the parameter `cls_model_dir` specifies the path to angle classification inference model and the parameter `rec_model_dir` specifies the path to identify the inference model. The parameter `use_angle_cls` is used to control whether to enable the angle classification model. The parameter `use_mp` specifies whether to use multi-process to infer `total_process_num` specifies process number when using multi-process. The parameter . The visualized recognition results are saved to the `./inference_results` folder by default.
```
```shell
# use direction classifier
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/00018069.jpg" --det_model_dir="./inference/det_db/" --cls_model_dir="./inference/cls/" --rec_model_dir="./inference/rec_crnn/" --use_angle_cls=true
# not use use direction classifier
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/00018069.jpg" --det_model_dir="./inference/det_db/" --rec_model_dir="./inference/rec_crnn/"
# use multi-process
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/00018069.jpg" --det_model_dir="./inference/det_db/" --rec_model_dir="./inference/rec_crnn/" --use_angle_cls=false --use_mp=True --total_process_num=6
```
```
After executing the command, the recognition result image is as follows:
......
......@@ -102,27 +102,16 @@ python3 generate_multi_language_configs.py -l it \
| german_mobile_v2.0_rec |Lightweight model for German recognition|[rec_german_lite_train.yml](../../configs/rec/multi_language/rec_german_lite_train.yml)|2.65M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/german_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/german_mobile_v2.0_rec_train.tar) |
| korean_mobile_v2.0_rec |Lightweight model for Korean recognition|[rec_korean_lite_train.yml](../../configs/rec/multi_language/rec_korean_lite_train.yml)|3.9M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/korean_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/korean_mobile_v2.0_rec_train.tar) |
| japan_mobile_v2.0_rec |Lightweight model for Japanese recognition|[rec_japan_lite_train.yml](../../configs/rec/multi_language/rec_japan_lite_train.yml)|4.23M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/japan_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/japan_mobile_v2.0_rec_train.tar) |
| it_mobile_v2.0_rec |Lightweight model for Italian recognition|rec_it_lite_train.yml|2.53M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/it_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/it_mobile_v2.0_rec_train.tar) |
| xi_mobile_v2.0_rec |Lightweight model for Spanish recognition|rec_xi_lite_train.yml|2.53M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/xi_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/xi_mobile_v2.0_rec_train.tar) |
| pu_mobile_v2.0_rec |Lightweight model for Portuguese recognition|rec_pu_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/pu_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/pu_mobile_v2.0_rec_train.tar) |
| ru_mobile_v2.0_rec |Lightweight model for Russia recognition|rec_ru_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ru_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ru_mobile_v2.0_rec_train.tar) |
| ar_mobile_v2.0_rec |Lightweight model for Arabic recognition|rec_ar_lite_train.yml|2.53M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ar_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ar_mobile_v2.0_rec_train.tar) |
| hi_mobile_v2.0_rec |Lightweight model for Hindi recognition|rec_hi_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/hi_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/hi_mobile_v2.0_rec_train.tar) |
| chinese_cht_mobile_v2.0_rec |Lightweight model for chinese traditional recognition|rec_chinese_cht_lite_train.yml|5.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/chinese_cht_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/chinese_cht_mobile_v2.0_rec_train.tar) |
| ug_mobile_v2.0_rec |Lightweight model for Uyghur recognition|rec_ug_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ug_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ug_mobile_v2.0_rec_train.tar) |
| fa_mobile_v2.0_rec |Lightweight model for Persian recognition|rec_fa_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/fa_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/fa_mobile_v2.0_rec_train.tar) |
| ur_mobile_v2.0_rec |Lightweight model for Urdu recognition|rec_ur_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ur_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ur_mobile_v2.0_rec_train.tar) |
| rs_mobile_v2.0_rec |Lightweight model for Serbian(latin) recognition|rec_rs_lite_train.yml|2.53M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/rs_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/rs_mobile_v2.0_rec_train.tar) |
| oc_mobile_v2.0_rec |Lightweight model for Occitan recognition|rec_oc_lite_train.yml|2.53M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/oc_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/oc_mobile_v2.0_rec_train.tar) |
| mr_mobile_v2.0_rec |Lightweight model for Marathi recognition|rec_mr_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/mr_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/mr_mobile_v2.0_rec_train.tar) |
| ne_mobile_v2.0_rec |Lightweight model for Nepali recognition|rec_ne_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ne_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ne_mobile_v2.0_rec_train.tar) |
| rsc_mobile_v2.0_rec |Lightweight model for Serbian(cyrillic) recognition|rec_rsc_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/rsc_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/rsc_mobile_v2.0_rec_train.tar) |
| bg_mobile_v2.0_rec |Lightweight model for Bulgarian recognition|rec_bg_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/bg_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/bg_mobile_v2.0_rec_train.tar) |
| uk_mobile_v2.0_rec |Lightweight model for Ukranian recognition|rec_uk_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/uk_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/uk_mobile_v2.0_rec_train.tar) |
| be_mobile_v2.0_rec |Lightweight model for Belarusian recognition|rec_be_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/be_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/be_mobile_v2.0_rec_train.tar) |
| chinese_cht_mobile_v2.0_rec |Lightweight model for chinese cht recognition|rec_chinese_cht_lite_train.yml|5.63M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/chinese_cht_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/chinese_cht_mobile_v2.0_rec_train.tar) |
| te_mobile_v2.0_rec |Lightweight model for Telugu recognition|rec_te_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/te_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/te_mobile_v2.0_rec_train.tar) |
| ka_mobile_v2.0_rec |Lightweight model for Kannada recognition|rec_ka_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ka_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ka_mobile_v2.0_rec_train.tar) |
| ta_mobile_v2.0_rec |Lightweight model for Tamil recognition|rec_ta_lite_train.yml|2.63M|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ta_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ta_mobile_v2.0_rec_train.tar) |
| latin_mobile_v2.0_rec | Lightweight model for latin recognition | [rec_latin_lite_train.yml](../../configs/rec/multi_language/rec_latin_lite_train.yml) |2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/latin_ppocr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/latin_ppocr_mobile_v2.0_rec_train.tar) |
| arabic_mobile_v2.0_rec | Lightweight model for arabic recognition | [rec_arabic_lite_train.yml](../../configs/rec/multi_language/rec_arabic_lite_train.yml) |2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/arabic_ppocr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/arabic_ppocr_mobile_v2.0_rec_train.tar) |
| cyrillic_mobile_v2.0_rec | Lightweight model for cyrillic recognition | [rec_cyrillic_lite_train.yml](../../configs/rec/multi_language/rec_cyrillic_lite_train.yml) |2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/cyrillic_ppocr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/cyrillic_ppocr_mobile_v2.0_rec_train.tar) |
| devanagari_mobile_v2.0_rec | Lightweight model for devanagari recognition | [rec_devanagari_lite_train.yml](../../configs/rec/multi_language/rec_devanagari_lite_train.yml) |2.6M|[推理模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/devanagari_ppocr_mobile_v2.0_rec_infer.tar) / [训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/devanagari_ppocr_mobile_v2.0_rec_train.tar) |
For more supported languages, please refer to : [Multi-language model](./multi_languages_en.md)
<a name="Angle"></a>
......
# Multi-language model
**Recent Update**
- 2021.4.9 supports the detection and recognition of 80 languages
- 2021.4.9 supports **lightweight high-precision** English model detection and recognition
PaddleOCR aims to create a rich, leading, and practical OCR tool library, which not only provides
Chinese and English models in general scenarios, but also provides models specifically trained
in English scenarios. And multilingual models covering [80 languages](#language_abbreviations).
Among them, the English model supports the detection and recognition of uppercase and lowercase
letters and common punctuation, and the recognition of space characters is optimized:
<div align="center">
<img src="../imgs_results/multi_lang/img_12.jpg" width="900" height="300">
</div>
The multilingual models cover Latin, Arabic, Traditional Chinese, Korean, Japanese, etc.:
<div align="center">
<img src="../imgs_results/multi_lang/japan_2.jpg" width="600" height="300">
<img src="../imgs_results/multi_lang/french_0.jpg" width="300" height="300">
<img src="../imgs_results/multi_lang/korean_0.jpg" width="500" height="300">
<img src="../imgs_results/multi_lang/arabic_0.jpg" width="300" height="300">
</div>
This document will briefly introduce how to use the multilingual model.
- [1 Installation](#Install)
- [1.1 paddle installation](#paddleinstallation)
- [1.2 paddleocr package installation](#paddleocr_package_install)
- [2 Quick Use](#Quick_Use)
- [2.1 Command line operation](#Command_line_operation)
- [2.2 python script running](#python_Script_running)
- [3 Custom Training](#Custom_Training)
- [4 Inference and Deployment](#inference)
- [4 Supported languages and abbreviations](#language_abbreviations)
<a name="Install"></a>
## 1 Installation
<a name="paddle_install"></a>
### 1.1 paddle installation
```
# cpu
pip install paddlepaddle
# gpu
pip install paddlepaddle-gpu
```
<a name="paddleocr_package_install"></a>
### 1.2 paddleocr package installation
pip install
```
pip install "paddleocr>=2.0.6" # 2.0.6 version is recommended
```
Build and install locally
```
python3 setup.py bdist_wheel
pip3 install dist/paddleocr-x.x.x-py3-none-any.whl # x.x.x is the version number of paddleocr
```
<a name="Quick_use"></a>
## 2 Quick use
<a name="Command_line_operation"></a>
### 2.1 Command line operation
View help information
```
paddleocr -h
```
* Whole image prediction (detection + recognition)
Paddleocr currently supports 80 languages, which can be switched by modifying the --lang parameter.
The specific supported [language] (#language_abbreviations) can be viewed in the table.
``` bash
paddleocr --image_dir doc/imgs_en/254.jpg --lang=en
```
<div align="center">
<img src="../imgs_en/254.jpg" width="300" height="600">
<img src="../imgs_results/multi_lang/img_02.jpg" width="600" height="600">
</div>
The result is a list, each item contains a text box, text and recognition confidence
```text
[('PHO CAPITAL', 0.95723116), [[66.0, 50.0], [327.0, 44.0], [327.0, 76.0], [67.0, 82.0]]]
[('107 State Street', 0.96311164), [[72.0, 90.0], [451.0, 84.0], [452.0, 116.0], [73.0, 121.0]]]
[('Montpelier Vermont', 0.97389287), [[69.0, 132.0], [501.0, 126.0], [501.0, 158.0], [70.0, 164.0]]]
[('8022256183', 0.99810505), [[71.0, 175.0], [363.0, 170.0], [364.0, 202.0], [72.0, 207.0]]]
[('REG 07-24-201706:59 PM', 0.93537045), [[73.0, 299.0], [653.0, 281.0], [654.0, 318.0], [74.0, 336.0]]]
[('045555', 0.99346405), [[509.0, 331.0], [651.0, 325.0], [652.0, 356.0], [511.0, 362.0]]]
[('CT1', 0.9988654), [[535.0, 367.0], [654.0, 367.0], [654.0, 406.0], [535.0, 406.0]]]
......
```
* Recognition
```bash
paddleocr --image_dir doc/imgs_words_en/word_308.png --det false --lang=en
```
![](https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/release/2.1/doc/imgs_words_en/word_308.png)
The result is a tuple, which returns the recognition result and recognition confidence
```text
(0.99879867, 'LITTLE')
```
* Detection
```
paddleocr --image_dir PaddleOCR/doc/imgs/11.jpg --rec false
```
The result is a list, each item contains only text boxes
```
[[26.0, 457.0], [137.0, 457.0], [137.0, 477.0], [26.0, 477.0]]
[[25.0, 425.0], [372.0, 425.0], [372.0, 448.0], [25.0, 448.0]]
[[128.0, 397.0], [273.0, 397.0], [273.0, 414.0], [128.0, 414.0]]
......
```
<a name="python_script_running"></a>
### 2.2 python script running
ppocr also supports running in python scripts for easy embedding in your own code:
* Whole image prediction (detection + recognition)
```
from paddleocr import PaddleOCR, draw_ocr
# Also switch the language by modifying the lang parameter
ocr = PaddleOCR(lang="korean") # The model file will be downloaded automatically when executed for the first time
img_path ='doc/imgs/korean_1.jpg'
result = ocr.ocr(img_path)
# Recognition and detection can be performed separately through parameter control
# result = ocr.ocr(img_path, det=False) Only perform recognition
# result = ocr.ocr(img_path, rec=False) Only perform detection
# Print detection frame and recognition result
for line in result:
print(line)
# Visualization
from PIL import Image
image = Image.open(img_path).convert('RGB')
boxes = [line[0] for line in result]
txts = [line[1][0] for line in result]
scores = [line[1][1] for line in result]
im_show = draw_ocr(image, boxes, txts, scores, font_path='/path/to/PaddleOCR/doc/fonts/korean.ttf')
im_show = Image.fromarray(im_show)
im_show.save('result.jpg')
```
Visualization of results:
![](https://raw.githubusercontent.com/PaddlePaddle/PaddleOCR/release/2.1/doc/imgs_results/korean.jpg)
ppocr also supports direction classification. For more usage methods, please refer to: [whl package instructions](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.0/doc/doc_ch/whl.md).
<a name="Custom_training"></a>
## 3 Custom training
ppocr supports using your own data for custom training or finetune, where the recognition model can refer to [French configuration file](../../configs/rec/multi_language/rec_french_lite_train.yml)
Modify the training data path, dictionary and other parameters.
For specific data preparation and training process, please refer to: [Text Detection](../doc_en/detection_en.md), [Text Recognition](../doc_en/recognition_en.md), more functions such as predictive deployment,
For functions such as data annotation, you can read the complete [Document Tutorial](../../README.md).
<a name="inference"></a>
## 4 Inference and Deployment
In addition to installing the whl package for quick forecasting,
ppocr also provides a variety of forecasting deployment methods.
If necessary, you can read related documents:
- [Python Inference](./inference_en.md)
- [C++ Inference](../../deploy/cpp_infer/readme_en.md)
- [Serving](../../deploy/hubserving/readme_en.md)
- [Mobile](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/lite/readme_en.md)
- [Benchmark](./benchmark_en.md)
<a name="language_abbreviations"></a>
## 5 Support languages and abbreviations
| Language | Abbreviation | | Language | Abbreviation |
| --- | --- | --- | --- | --- |
|chinese and english|ch| |Arabic|ar|
|english|en| |Hindi|hi|
|french|fr| |Uyghur|ug|
|german|german| |Persian|fa|
|japan|japan| |Urdu|ur|
|korean|korean| | Serbian(latin) |rs_latin|
|chinese traditional |ch_tra| |Occitan |oc|
| Italian |it| |Marathi|mr|
|Spanish |es| |Nepali|ne|
| Portuguese|pt| |Serbian(cyrillic)|rs_cyrillic|
|Russia|ru||Bulgarian |bg|
|Ukranian|uk| |Estonian |et|
|Belarusian|be| |Irish |ga|
|Telugu |te| |Croatian |hr|
|Saudi Arabia|sa| |Hungarian |hu|
|Tamil |ta| |Indonesian|id|
|Afrikaans |af| |Icelandic|is|
|Azerbaijani |az||Kurdish|ku|
|Bosnian|bs| |Lithuanian |lt|
|Czech|cs| |Latvian |lv|
|Welsh |cy| |Maori|mi|
|Danish|da| |Malay|ms|
|Maltese |mt| |Adyghe |ady|
|Dutch |nl| |Kabardian |kbd|
|Norwegian |no| |Avar |ava|
|Polish |pl| |Dargwa |dar|
|Romanian |ro| |Ingush |inh|
|Slovak |sk| |Lak |lbe|
|Slovenian |sl| |Lezghian |lez|
|Albanian |sq| |Tabassaran |tab|
|Swedish |sv| |Bihari |bh|
|Swahili |sw| |Maithili |mai|
|Tagalog |tl| |Angika |ang|
|Turkish |tr| |Bhojpuri |bho|
|Uzbek |uz| |Magahi |mah|
|Vietnamese |vi| |Nagpur |sck|
|Mongolian |mn| |Newari |new|
|Abaza |abq| |Goan Konkani|gom|
# End-to-end OCR Algorithm-PGNet
- [1. Brief Introduction](#Brief_Introduction)
- [2. Environment Configuration](#Environment_Configuration)
- [3. Quick Use](#Quick_Use)
- [4. Model Training,Evaluation And Inference](#Model_Training_Evaluation_And_Inference)
<a name="Brief_Introduction"></a>
## 1. Brief Introduction
OCR algorithm can be divided into two-stage algorithm and end-to-end algorithm. The two-stage OCR algorithm is generally divided into two parts, text detection and text recognition algorithm. The text detection algorithm gets the detection box of the text line from the image, and then the recognition algorithm identifies the content of the text box. The end-to-end OCR algorithm can complete text detection and recognition in one algorithm. Its basic idea is to design a model with both detection unit and recognition module, share the CNN features of both and train them together. Because one algorithm can complete character recognition, the end-to-end model is smaller and faster.
### Introduction Of PGNet Algorithm
In recent years, the end-to-end OCR algorithm has been well developed, including MaskTextSpotter series, TextSnake, TextDragon, PGNet series and so on. Among these algorithms, PGNet algorithm has the advantages that other algorithms do not
- Pgnet loss is designed to guide training, and no character-level annotations is needed
- NMS and ROI related operations are not needed, It can accelerate the prediction
- The reading order prediction module is proposed
- A graph based modification module (GRM) is proposed to further improve the performance of model recognition
- Higher accuracy and faster prediction speed
For details of PGNet algorithm, please refer to [paper](https://www.aaai.org/AAAI21Papers/AAAI-2885.WangP.pdf) ,The schematic diagram of the algorithm is as follows:
![](../pgnet_framework.png)
After feature extraction, the input image is sent to four branches: TBO module for text edge offset prediction, TCL module for text centerline prediction, TDO module for text direction offset prediction, and TCC module for text character classification graph prediction.
The output of TBO and TCL can get text detection results after post-processing, and TCL, TDO and TCC are responsible for text recognition.
The results of detection and recognition are as follows:
![](../imgs_results/e2e_res_img293_pgnet.png)
![](../imgs_results/e2e_res_img295_pgnet.png)
### Performance
####Test set: Total Text
####Test environment: NVIDIA Tesla V100-SXM2-16GB
|PGNetA|det_precision|det_recall|det_f_score|e2e_precision|e2e_recall|e2e_f_score|FPS|download|
| --- | --- | --- | --- | --- | --- | --- | --- | --- |
|Paper|85.30|86.80|86.1|-|-|61.7|38.20 (size=640)|-|
|Ours|87.03|82.48|84.69|61.71|58.43|60.03|48.73 (size=768)|[download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/en_server_pgnetA.tar)|
*note:PGNet in PaddleOCR optimizes the prediction speed, and can significantly improve the end-to-end prediction speed within the acceptable range of accuracy reduction*
<a name="Environment_Configuration"></a>
## 2. Environment Configuration
Please refer to [Quick Installation](./installation_en.md) Configure the PaddleOCR running environment.
<a name="Quick_Use"></a>
## 3. Quick Use
### inference model download
This section takes the trained end-to-end model as an example to quickly use the model prediction. First, download the trained end-to-end inference model [download address](https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/e2e_server_pgnetA_infer.tar)
```
mkdir inference && cd inference
# Download the English end-to-end model and unzip it
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/e2e_server_pgnetA_infer.tar && tar xf e2e_server_pgnetA_infer.tar
```
* In Windows environment, if 'wget' is not installed, the link can be copied to the browser when downloading the model, and decompressed and placed in the corresponding directory
After decompression, there should be the following file structure:
```
├── e2e_server_pgnetA_infer
│ ├── inference.pdiparams
│ ├── inference.pdiparams.info
│ └── inference.pdmodel
```
### Single image or image set prediction
```bash
# Prediction single image specified by image_dir
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/img623.jpg" --e2e_model_dir="./inference/e2e_server_pgnetA_infer/" --e2e_pgnet_polygon=True
# Prediction the collection of images specified by image_dir
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/" --e2e_model_dir="./inference/e2e_server_pgnetA_infer/" --e2e_pgnet_polygon=True
# If you want to use CPU for prediction, you need to set use_gpu parameter is false
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/img623.jpg" --e2e_model_dir="./inference/e2e_server_pgnetA_infer/" --e2e_pgnet_polygon=True --use_gpu=False
```
### Visualization results
The visualized end-to-end results are saved to the `./inference_results` folder by default, and the name of the result file is prefixed with 'e2e_res'. Examples of results are as follows:
![](../imgs_results/e2e_res_img623_pgnet.jpg)
<a name="Model_Training_Evaluation_And_Inference"></a>
## 4. Model Training,Evaluation And Inference
This section takes the totaltext dataset as an example to introduce the training, evaluation and testing of the end-to-end model in PaddleOCR.
### Data Preparation
Download and unzip [totaltext](https://paddleocr.bj.bcebos.com/dataset/total_text.tar) dataset to PaddleOCR/train_data/, dataset organization structure is as follow:
```
/PaddleOCR/train_data/total_text/train/
|- rgb/ # total_text training data of dataset
|- img11.png
| ...
|- train.txt # total_text training annotation of dataset
```
total_text.txt: the format of dimension file is as follows,the file name and annotation information are separated by "\t":
```
" Image file name Image annotation information encoded by json.dumps"
rgb/img11.jpg [{"transcription": "ASRAMA", "points": [[214.0, 325.0], [235.0, 308.0], [259.0, 296.0], [286.0, 291.0], [313.0, 295.0], [338.0, 305.0], [362.0, 320.0], [349.0, 347.0], [330.0, 337.0], [310.0, 329.0], [290.0, 324.0], [269.0, 328.0], [249.0, 336.0], [231.0, 346.0]]}, {...}]
```
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.
`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.
### Start Training
PGNet training is divided into two steps: Step 1: training on the synthetic data to get the pretrain_model, and the accuracy of the model is still low; step 2: loading the pretrain_model and training on the totaltext data set; for fast training, we directly provide the pre training model of step 1[download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/train_step1.tar).
```shell
cd PaddleOCR/
download step1 pretrain_models
wget -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/train_step1.tar
You can get the following file format
./pretrain_models/train_step1/
└─ best_accuracy.pdopt
└─ best_accuracy.states
└─ best_accuracy.pdparams
```
*If CPU version installed, please set the parameter `use_gpu` to `false` in the configuration.*
```shell
# single GPU training
python3 tools/train.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.pretrained_model=./pretrain_models/train_step1/best_accuracy Global.load_static_weights=False
# multi-GPU training
# Set the GPU ID used by the '--gpus' parameter.
python3 -m paddle.distributed.launch --gpus '0,1,2,3' tools/train.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.pretrained_model=./pretrain_models/train_step1/best_accuracy Global.load_static_weights=False
```
In the above instruction, use `-c` to select the training to use the `configs/e2e/e2e_r50_vd_pg.yml` configuration file.
For a detailed explanation of the configuration file, please refer to [config](./config_en.md).
You can also use `-o` to change the training parameters without modifying the yml file. For example, adjust the training learning rate to 0.0001
```shell
python3 tools/train.py -c configs/e2e/e2e_r50_vd_pg.yml -o Optimizer.base_lr=0.0001
```
#### Load trained model and continue 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.
```shell
python3 tools/train.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.checkpoints=./your/trained/model
```
**Note**: The priority of `Global.checkpoints` is higher than that of `Global.pretrain_weights`, that is, when two parameters are specified at the same time, the model specified by `Global.checkpoints` will be loaded first. If the model path specified by `Global.checkpoints` is wrong, the one specified by `Global.pretrain_weights` will be loaded.
PaddleOCR calculates three indicators for evaluating performance of OCR end-to-end task: Precision, Recall, and Hmean.
Run the following code to calculate the evaluation indicators. The result will be saved in the test result file specified by `save_res_path` in the configuration file `e2e_r50_vd_pg.yml`
When evaluating, set post-processing parameters `max_side_len=768`. If you use different datasets, different models for training.
The model parameters during training are saved in the `Global.save_model_dir` directory by default. When evaluating indicators, you need to set `Global.checkpoints` to point to the saved parameter file.
```shell
python3 tools/eval.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.checkpoints="{path/to/weights}/best_accuracy"
```
### Model Test
Test the end-to-end result on a single image:
```shell
python3 tools/infer_e2e.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.infer_img="./doc/imgs_en/img_10.jpg" Global.pretrained_model="./output/e2e_pgnet/best_accuracy" Global.load_static_weights=false
```
Test the end-to-end result on all images in the folder:
```shell
python3 tools/infer_e2e.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.infer_img="./doc/imgs_en/" Global.pretrained_model="./output/e2e_pgnet/best_accuracy" Global.load_static_weights=false
```
### Model inference
#### (1).Quadrangle text detection model (ICDAR2015)
First, convert the model saved in the PGNet end-to-end training process into an inference model. In the first stage of training based on composite dataset, the model of English data set training is taken as an example[model download link](https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/en_server_pgnetA.tar), you can use the following command to convert:
```
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/pgnet/en_server_pgnetA.tar && tar xf en_server_pgnetA.tar
python3 tools/export_model.py -c configs/e2e/e2e_r50_vd_pg.yml -o Global.pretrained_model=./en_server_pgnetA/best_accuracy Global.load_static_weights=False Global.save_inference_dir=./inference/e2e
```
**For PGNet quadrangle end-to-end model inference, you need to set the parameter `--e2e_algorithm="PGNet"`**, run the following command:
```
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/img_10.jpg" --e2e_model_dir="./inference/e2e/" --e2e_pgnet_polygon=False
```
The visualized text detection results are saved to the `./inference_results` folder by default, and the name of the result file is prefixed with 'e2e_res'. Examples of results are as follows:
![](../imgs_results/e2e_res_img_10_pgnet.jpg)
#### (2). Curved text detection model (Total-Text)
For the curved text example, we use the same model as the quadrilateral
**For PGNet end-to-end curved text detection model inference, you need to set the parameter `--e2e_algorithm="PGNet"` and `--e2e_pgnet_polygon=True`**, run the following command:
```
python3 tools/infer/predict_e2e.py --e2e_algorithm="PGNet" --image_dir="./doc/imgs_en/img623.jpg" --e2e_model_dir="./inference/e2e/" --e2e_pgnet_polygon=True
```
The visualized text detection results are saved to the `./inference_results` folder by default, and the name of the result file is prefixed with 'e2e_res'. Examples of results are as follows:
![](../imgs_results/e2e_res_img623_pgnet.jpg)
......@@ -131,7 +131,7 @@ PaddleOCR has built-in dictionaries, which can be used on demand.
`ppocr/utils/dict/german_dict.txt` is a German dictionary with 131 characters
`ppocr/utils/dict/en_dict.txt` is a English dictionary with 63 characters
`ppocr/utils/en_dict.txt` is a English dictionary with 96 characters
The current multi-language model is still in the demo stage and will continue to optimize the model and add languages. **You are very welcome to provide us with dictionaries and fonts in other languages**,
......@@ -279,7 +279,7 @@ Eval:
<a name="Multi_language"></a>
#### 2.3 Multi-language
PaddleOCR currently supports 26 (except Chinese) language recognition. A multi-language configuration file template is
PaddleOCR currently supports 80 (except Chinese) language recognition. A multi-language configuration file template is
provided under the path `configs/rec/multi_languages`: [rec_multi_language_lite_train.yml](../../configs/rec/multi_language/rec_multi_language_lite_train.yml)
There are two ways to create the required configuration file::
......@@ -368,27 +368,12 @@ Currently, the multi-language algorithms supported by PaddleOCR are:
| rec_ger_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | German | german |
| rec_japan_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Japanese | japan |
| rec_korean_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Korean | korean |
| rec_it_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Italian | it |
| rec_xi_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Spanish | xi |
| rec_pu_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Portuguese | pu |
| rec_ru_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Russia | ru |
| rec_ar_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Arabic | ar |
| rec_hi_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Hindi | hi |
| rec_ug_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Uyghur | ug |
| rec_fa_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Persian(Farsi) | fa |
| rec_ur_ite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Urdu | ur |
| rec_rs_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Serbian(latin) | rs |
| rec_oc_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Occitan | oc |
| rec_mr_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Marathi | mr |
| rec_ne_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Nepali | ne |
| rec_rsc_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Serbian(cyrillic) | rsc |
| rec_bg_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Bulgarian | bg |
| rec_uk_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Ukranian | uk |
| rec_be_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Belarusian | be |
| rec_te_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Telugu | te |
| rec_ka_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Kannada | ka |
| rec_ta_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Tamil | ta |
| rec_latin_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | Latin | latin |
| rec_arabic_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | arabic | ar |
| rec_cyrillic_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | cyrillic | cyrillic |
| rec_devanagari_lite_train.yml | CRNN | Mobilenet_v3 small 0.5 | None | BiLSTM | ctc | devanagari | devanagari |
For more supported languages, please refer to : [Multi-language model](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/doc/doc_en/multi_languages_en.md#4-support-languages-and-abbreviations)
The multi-language model training method is the same as the Chinese model. The training data set is 100w synthetic data. A small amount of fonts and test data can be downloaded on [Baidu Netdisk](https://pan.baidu.com/s/1bS_u207Rm7YbY33wOECKDA),Extraction code:frgi.
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment