提交 2547ab09 编写于 作者: qq_25193841's avatar qq_25193841

Merge remote-tracking branch 'upstream/dygraph' into dygraph

......@@ -29,3 +29,5 @@ paddleocr.egg-info/
/deploy/android_demo/app/PaddleLite/
/deploy/android_demo/app/.cxx/
/deploy/android_demo/app/cache/
test_tipc/web/models/
test_tipc/web/node_modules/
......@@ -90,7 +90,7 @@ Mobile DEMO experience (based on EasyEdge and Paddle-Lite, supports iOS and Andr
| Model introduction | Model name | Recommended scene | Detection model | Direction classifier | Recognition model |
| ------------------------------------------------------------ | ---------------------------- | ----------------- | ------------------------------------------------------------ | ------------------------------------------------------------ | ------------------------------------------------------------ |
| Chinese and English ultra-lightweight PP-OCRv2 model(11.6M) | ch_PP-OCRv2_xx |Mobile & Server|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_distill_train.tar)| [inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv2/ch/ch_PP-OCRv2_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_train.tar)|
| Chinese and English ultra-lightweight PP-OCRv2 model(11.6M) | ch_PP-OCRv2_xx |Mobile & Server|[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_distill_train.tar)| [inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) |[inference model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_train.tar)|
| Chinese and English ultra-lightweight PP-OCR model (9.4M) | ch_ppocr_mobile_v2.0_xx | Mobile & server |[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar)|[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar) |[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_train.tar) |
| Chinese and English general PP-OCR model (143.4M) | ch_ppocr_server_v2.0_xx | Server |[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_det_train.tar) |[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_traingit.tar) |[inference model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_infer.tar) / [trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_train.tar) |
......
......@@ -21,6 +21,7 @@ Architecture:
model_type: det
Models:
Teacher:
pretrained: ./pretrain_models/ch_ppocr_server_v2.0_det_train/best_accuracy
freeze_params: true
return_all_feats: false
model_type: det
......@@ -36,6 +37,7 @@ Architecture:
name: DBHead
k: 50
Student:
pretrained:
freeze_params: false
return_all_feats: false
model_type: det
......@@ -52,6 +54,7 @@ Architecture:
name: DBHead
k: 50
Student2:
pretrained:
freeze_params: false
return_all_feats: false
model_type: det
......
......@@ -18,6 +18,7 @@ Global:
Architecture:
name: DistillationModel
algorithm: Distillation
model_type: det
Models:
Student:
pretrained: ./pretrain_models/MobileNetV3_large_x0_5_pretrained
......
......@@ -18,6 +18,7 @@ Global:
Architecture:
name: DistillationModel
algorithm: Distillation
model_type: det
Models:
Student:
pretrained: ./pretrain_models/MobileNetV3_large_x0_5_pretrained
......
## 数据合成工具
# 数据合成工具
除了开源数据,用户还可使用合成工具自行合成。这里整理了常用的数据合成工具,持续更新中,欢迎各位小伙伴贡献工具~
- [text_renderer](https://github.com/Sanster/text_renderer)
- [SynthText](https://github.com/ankush-me/SynthText)
......@@ -6,3 +6,4 @@
- [TextRecognitionDataGenerator](https://github.com/Belval/TextRecognitionDataGenerator)
- [SynthText3D](https://github.com/MhLiao/SynthText3D)
- [UnrealText](https://github.com/Jyouhou/UnrealText/)
- [SynthTIGER](https://github.com/clovaai/synthtiger)
\ No newline at end of file
......@@ -281,4 +281,274 @@ paddle.save(s_params, "ch_PP-OCRv2_rec_train/student.pdparams")
### 2.2 检测配置文件解析
* coming soon!
检测模型蒸馏的配置文件在PaddleOCR/configs/det/ch_PP-OCRv2/目录下,包含三个蒸馏配置文件:
- ch_PP-OCRv2_det_cml.yml,采用cml蒸馏,采用一个大模型蒸馏两个小模型,且两个小模型互相学习的方法
- ch_PP-OCRv2_det_dml.yml,采用DML的蒸馏,两个Student模型互蒸馏的方法
- ch_PP-OCRv2_det_distill.yml,采用Teacher大模型蒸馏小模型Student的方法
#### 2.2.1 模型结构
知识蒸馏任务中,模型结构配置如下所示:
```
Architecture:
name: DistillationModel # 结构名称,蒸馏任务中,为DistillationModel,用于构建对应的结构
algorithm: Distillation # 算法名称
Models: # 模型,包含子网络的配置信息
Student: # 子网络名称,至少需要包含`pretrained`与`freeze_params`信息,其他的参数为子网络的构造参数
pretrained: ./pretrain_models/MobileNetV3_large_x0_5_pretrained
freeze_params: false # 是否需要固定参数
return_all_feats: false # 子网络的参数,表示是否需要返回所有的features,如果为False,则只返回最后的输出
model_type: det
algorithm: DB
Backbone:
name: MobileNetV3
scale: 0.5
model_name: large
disable_se: True
Neck:
name: DBFPN
out_channels: 96
Head:
name: DBHead
k: 50
Teacher: # 另外一个子网络,这里给的是普通大模型蒸小模型的蒸馏示例,
pretrained: ./pretrain_models/ch_ppocr_server_v2.0_det_train/best_accuracy
freeze_params: true # Teacher模型是训练好的,不需要参与训练,freeze_params设置为True
return_all_feats: false
model_type: det
algorithm: DB
Transform:
Backbone:
name: ResNet
layers: 18
Neck:
name: DBFPN
out_channels: 256
Head:
name: DBHead
k: 50
```
如果是采用DML,即两个小模型互相学习的方法,上述配置文件里的Teacher网络结构需要设置为Student模型一样的配置,具体参考配置文件[ch_PP-OCRv2_det_dml.yml](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.4/configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_dml.yml)

下面介绍[ch_PP-OCRv2_det_cml.yml](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.4/configs/det/ch_PP-OCRv2/ch_PP-OCRv2_det_cml.yml)的配置文件参数:
```
Architecture:
name: DistillationModel
algorithm: Distillation
model_type: det
Models:
Teacher: # CML蒸馏的Teacher模型配置
pretrained: ./pretrain_models/ch_ppocr_server_v2.0_det_train/best_accuracy
freeze_params: true # Teacher 不训练
return_all_feats: false
model_type: det
algorithm: DB
Transform:
Backbone:
name: ResNet
layers: 18
Neck:
name: DBFPN
out_channels: 256
Head:
name: DBHead
k: 50
Student: # CML蒸馏的Student模型配置
pretrained: ./pretrain_models/MobileNetV3_large_x0_5_pretrained
freeze_params: false
return_all_feats: false
model_type: det
algorithm: DB
Backbone:
name: MobileNetV3
scale: 0.5
model_name: large
disable_se: True
Neck:
name: DBFPN
out_channels: 96
Head:
name: DBHead
k: 50
Student2: # CML蒸馏的Student2模型配置
pretrained: ./pretrain_models/MobileNetV3_large_x0_5_pretrained
freeze_params: false
return_all_feats: false
model_type: det
algorithm: DB
Transform:
Backbone:
name: MobileNetV3
scale: 0.5
model_name: large
disable_se: True
Neck:
name: DBFPN
out_channels: 96
Head:
name: DBHead
k: 50
```
蒸馏模型`DistillationModel`类的具体实现代码可以参考[distillation_model.py](../../ppocr/modeling/architectures/distillation_model.py)
最终模型`forward`输出为一个字典,key为所有的子网络名称,例如这里为`Student``Teacher`,value为对应子网络的输出,可以为`Tensor`(只返回该网络的最后一层)和`dict`(也返回了中间的特征信息)。
在蒸馏任务中,为了方便添加蒸馏损失函数,每个网络的输出保存为`dict`,其中包含子模块输出。每个子网络的输出结果均为`dict`,key包含`backbone_out`,`neck_out`, `head_out``value`为对应模块的tensor,最终对于上述配置文件,`DistillationModel`的输出格式如下。
```json
{
"Teacher": {
"backbone_out": tensor,
"neck_out": tensor,
"head_out": tensor,
},
"Student": {
"backbone_out": tensor,
"neck_out": tensor,
"head_out": tensor,
}
}
```
#### 2.1.2 损失函数
知识蒸馏任务中,检测ch_PP-OCRv2_det_distill.yml蒸馏损失函数配置如下所示。
```yaml
Loss:
name: CombinedLoss # 损失函数名称,基于改名称,构建用于损失函数的类
loss_config_list: # 损失函数配置文件列表,为CombinedLoss的必备函数
- DistillationDilaDBLoss: # 基于蒸馏的DB损失函数,继承自标准的DBloss
weight: 1.0 # 损失函数的权重,loss_config_list中,每个损失函数的配置都必须包含该字段
model_name_pairs: # 对于蒸馏模型的预测结果,提取这两个子网络的输出,计算Teacher模型和Student模型输出的loss
- ["Student", "Teacher"]
key: maps # 取子网络输出dict中,该key对应的tensor
balance_loss: true # 以下几个参数为标准DBloss的配置参数
main_loss_type: DiceLoss
alpha: 5
beta: 10
ohem_ratio: 3
- DistillationDBLoss: # 基于蒸馏的DB损失函数,继承自标准的DBloss,用于计算Student和GT之间的loss
weight: 1.0
model_name_list: ["Student"] # 模型名字只有Student,表示计算Student和GT之间的loss
name: DBLoss
balance_loss: true
main_loss_type: DiceLoss
alpha: 5
beta: 10
ohem_ratio: 3
```
同理,检测ch_PP-OCRv2_det_cml.yml蒸馏损失函数配置如下所示。相比较于ch_PP-OCRv2_det_distill.yml的损失函数配置,cml蒸馏的损失函数配置做了3个改动:
```yaml
Loss:
name: CombinedLoss
loss_config_list:
- DistillationDilaDBLoss:
weight: 1.0
model_name_pairs:
- ["Student", "Teacher"]
- ["Student2", "Teacher"] # 改动1,计算两个Student和Teacher的损失
key: maps
balance_loss: true
main_loss_type: DiceLoss
alpha: 5
beta: 10
ohem_ratio: 3
- DistillationDMLLoss: # 改动2,增加计算两个Student之间的损失
model_name_pairs:
- ["Student", "Student2"]
maps_name: "thrink_maps"
weight: 1.0
# act: None
key: maps
- DistillationDBLoss:
weight: 1.0
model_name_list: ["Student", "Student2"] # 改动3,计算两个Student和GT之间的损失
balance_loss: true
main_loss_type: DiceLoss
alpha: 5
beta: 10
ohem_ratio: 3
```
关于`DistillationDilaDBLoss`更加具体的实现可以参考: [distillation_loss.py](https://github.com/PaddlePaddle/PaddleOCR/blob/release%2F2.4/ppocr/losses/distillation_loss.py#L185)。关于`DistillationDBLoss`等蒸馏损失函数更加具体的实现可以参考[distillation_loss.py](https://github.com/PaddlePaddle/PaddleOCR/blob/04c44974b13163450dfb6bd2c327863f8a194b3c/ppocr/losses/distillation_loss.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L148)
#### 2.1.3 后处理
知识蒸馏任务中,检测蒸馏后处理配置如下所示。
```yaml
PostProcess:
name: DistillationDBPostProcess # DB检测蒸馏任务的CTC解码后处理,继承自标准的DBPostProcess类
model_name: ["Student", "Student2", "Teacher"] # 对于蒸馏模型的预测结果,提取多个子网络的输出,进行解码,不需要后处理的网络可以不在model_name中设置
thresh: 0.3
box_thresh: 0.6
max_candidates: 1000
unclip_ratio: 1.5
```
以上述配置为例,最终会同时计算`Student``Student2``Teacher` 3个子网络的输出做后处理计算。同时,由于有多个输入,后处理返回的输出也有多个,
关于`DistillationDBPostProcess`更加具体的实现可以参考: [db_postprocess.py](../../ppocr/postprocess/db_postprocess.py#L195)
#### 2.1.4 蒸馏指标计算
知识蒸馏任务中,检测蒸馏指标计算配置如下所示。
```yaml
Metric:
name: DistillationMetric
base_metric_name: DetMetric
main_indicator: hmean
key: "Student"
```
由于蒸馏需要包含多个网络,甚至多个Student网络,在计算指标的时候只需要计算一个Student网络的指标即可,`key`字段设置为`Student`则表示只计算`Student`网络的精度。
#### 2.1.5 检测蒸馏模型finetune
检测蒸馏有三种方式:
- 采用ch_PP-OCRv2_det_distill.yml,Teacher模型设置为PaddleOCR提供的模型或者您训练好的大模型
- 采用ch_PP-OCRv2_det_cml.yml,采用cml蒸馏,同样Teacher模型设置为PaddleOCR提供的模型或者您训练好的大模型
- 采用ch_PP-OCRv2_det_dml.yml,采用DML的蒸馏,两个Student模型互蒸馏的方法,在PaddleOCR采用的数据集上大约有1.7%的精度提升。
在具体finetune时,需要在网络结构的`pretrained`参数中设置要加载的预训练模型。
在精度提升方面,cml的精度>dml的精度>distill蒸馏方法的精度。当数据量不足或者Teacher模型精度与Student精度相差不大的时候,这个结论或许会改变。
另外,由于PaddleOCR提供的蒸馏预训练模型包含了多个模型的参数,如果您希望提取Student模型的参数,可以参考如下代码:
```
# 下载蒸馏训练模型的参数
wget https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_distill_train.tar
```
```python
import paddle
# 加载预训练模型
all_params = paddle.load("ch_PP-OCRv2_det_distill_train/best_accuracy.pdparams")
# 查看权重参数的keys
print(all_params.keys())
# 学生模型的权重提取
s_params = {key[len("Student."):]: all_params[key] for key in all_params if "Student." in key}
# 查看学生模型权重参数的keys
print(s_params.keys())
# 保存
paddle.save(s_params, "ch_PP-OCRv2_det_distill_train/student.pdparams")
```
最终`Student`模型的参数将会保存在`ch_PP-OCRv2_det_distill_train/student.pdparams`中,用于模型的fine-tune。
......@@ -9,3 +9,4 @@ There are the commonly used data synthesis tools, which will be continuously upd
* [TextRecognitionDataGenerator](https://github.com/Belval/TextRecognitionDataGenerator)
* [SynthText3D](https://github.com/MhLiao/SynthText3D)
* [UnrealText](https://github.com/Jyouhou/UnrealText/)
* [SynthTIGER](https://github.com/clovaai/synthtiger)
\ No newline at end of file
{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"![](https://ai-studio-static-online.cdn.bcebos.com/72b2077605dd49b78f7f647d6821d10231f6bc52d7ed463da451a6a0bd1fc5ff)\n",
"*注:以上图片来自网络*\n",
"\n",
"# 1. OCR技术背景\n",
"## 1.1 OCR技术的应用场景\n",
"\n",
"* **<font color=red>OCR是什么</font>**\n",
"\n",
"OCR(Optical Character Recognition,光学字符识别)是计算机视觉重要方向之一。传统定义的OCR一般面向扫描文档类对象,现在我们常说的OCR一般指场景文字识别(Scene Text Recognition,STR),主要面向自然场景,如下图中所示的牌匾等各种自然场景可见的文字。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/c87c0e6f6c0a42cdbc552a4f973c1b0217c369194c1243558753896f3e66032c)\n",
"<center>图1 文档场景文字识别 VS. 自然场景文字识别</center>\n",
"\n",
"<br>\n",
"\n",
"* **<font color=red>OCR有哪些应用场景?</font>**\n",
"\n",
"OCR技术有着丰富的应用场景,一类典型的场景是日常生活中广泛应用的面向垂类的结构化文本识别,比如车牌识别、银行卡信息识别、身份证信息识别、火车票信息识别等等。这些小垂类的共同特点是格式固定,因此非常适合使用OCR技术进行自动化,可以极大的减轻人力成本,提升效率。\n",
"\n",
"这种面向垂类的结构化文本识别是目前ocr应用最广泛、并且技术相对较成熟的场景。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/56e0df91d0d34443aacb17c9a1c5c186608ee675092648a693503df7fe45e535)\n",
"<center>图2 OCR技术的应用场景</center>\n",
"\n",
"除了面向垂类的结构化文本识别,通用OCR技术也有广泛的应用,并且常常和其他技术结合完成多模态任务,例如在视频场景中,经常使用OCR技术进行字幕自动翻译、内容安全监控等等,或者与视觉特征相结合,完成视频理解、视频搜索等任务。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/ca2341a51eb242ee8e1afe121ce3ebbc87a113cef1b643ed9bba92d0c8ee4f0f)\n",
"<center>图3 多模态场景中的通用OCR</center>\n",
"\n",
"## 1.2 OCR技术挑战\n",
"OCR的技术难点可以分为算法层和应用层两方面。\n",
"\n",
"* **<font color=red>算法层</font>**\n",
"\n",
"OCR丰富的应用场景,决定了它会存在很多技术难点。这里给出了常见的8种问题:\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/a56831fbf0c449fe9156a893002cadfe110ccfea835b4d90854a7ce4b1df2a4f)\n",
"<center>图4 OCR算法层技术难点</center>\n",
"\n",
"这些问题给文本检测和文本识别都带来了巨大的技术挑战,可以看到,这些挑战主要都是面向自然场景,目前学术界的研究也主要聚焦在自然场景,OCR领域在学术上的常用数据集也都是自然场景。针对这些问题的研究很多,相对来说,识别比检测面临更大的挑战。\n",
"\n",
"* **<font color=red>应用层</font>**\n",
"\n",
"在实际应用中,尤其是在广泛的通用场景下,除了上一节总结的仿射变换、尺度问题、光照不足、拍摄模糊等算法层面的技术难点,OCR技术还面临两大落地难点:\n",
"1. **海量数据要求OCR能够实时处理。** OCR应用常对接海量数据,我们要求或希望数据能够得到实时处理,模型的速度做到实时是一个不小的挑战。\n",
"2. **端侧应用要求OCR模型足够轻量,识别速度足够快。** OCR应用常部署在移动端或嵌入式硬件,端侧OCR应用一般有两种模式:上传到服务器 vs. 端侧直接识别,考虑到上传到服务器的方式对网络有要求,实时性较低,并且请求量过大时服务器压力大,以及数据传输的安全性问题,我们希望能够直接在端侧完成OCR识别,而端侧的存储空间和计算能力有限,因此对OCR模型的大小和预测速度有很高的要求。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/5bafdc3da1614c41a95ae39a2c36632f95e2893031a64929b9f49d4a4985cd2d)\n",
"<center>图5 OCR应用层技术难点</center>\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"# 2. OCR前沿算法\n",
"\n",
"虽然OCR是一个相对具体的任务,但涉及了多方面的技术,包括文本检测、文本识别、端到端文本识别、文档分析等等。学术上关于OCR各项相关技术的研究层出不穷,下文将简要介绍OCR任务中的几种关键技术的相关工作。\n",
"\n",
"## 2.1 文本检测\n",
"\n",
"文本检测的任务是定位出输入图像中的文字区域。近年来学术界关于文本检测的研究非常丰富,一类方法将文本检测视为目标检测中的一个特定场景,基于通用目标检测算法进行改进适配,如TextBoxes[1]基于一阶段目标检测器SSD[2]算法,调整目标框使之适合极端长宽比的文本行,CTPN[3]则是基于Faster RCNN[4]架构改进而来。但是文本检测与目标检测在目标信息以及任务本身上仍存在一些区别,如文本一般长宽比较大,往往呈“条状”,文本行之间可能比较密集,弯曲文本等,因此又衍生了很多专用于文本检测的算法,如EAST[5]、PSENet[6]、DBNet[7]等等。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/548b50212935402abb2e671c158c204737c2c64b9464442a8f65192c8a31b44d\" width=\"500\"></center>\n",
"<center>图6 文本检测任务示例</center>\n",
"\n",
"<br>\n",
"\n",
"目前较为流行的文本检测算法可以大致分为**基于回归**和**基于分割**的两大类文本检测算法,也有一些算法将二者相结合。基于回归的算法借鉴通用物体检测算法,通过设定anchor回归检测框,或者直接做像素回归,这类方法对规则形状文本检测效果较好,但是对不规则形状的文本检测效果会相对差一些,比如CTPN[3]对水平文本的检测效果较好,但对倾斜、弯曲文本的检测效果较差,SegLink[8]对长文本比较好,但对分布稀疏的文本效果较差;基于分割的算法引入了Mask-RCNN[9],这类算法在各种场景、对各种形状文本的检测效果都可以达到一个更高的水平,但缺点就是后处理一般会比较复杂,因此常常存在速度问题,并且无法解决重叠文本的检测问题。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/4f4ea65578384900909efff93d0b7386e86ece144d8c4677b7bc94b4f0337cfb\" width=\"800\"></center>\n",
"<center>图7 文本检测算法概览</center>\n",
"\n",
"<br>\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/473ba28cd0274d568f90eb8ca9e78864d994f3ebffe6419cb638e193c607b7b3)|![](https://ai-studio-static-online.cdn.bcebos.com/e968807b3ed9493cab20f3be0d8dc07b0baf8b8cecb24ee99ccda9d3a241832a)|![](https://ai-studio-static-online.cdn.bcebos.com/53b9e85ce46645c08481d7d7377720f5eea5ac30e37e4e9c9930e1f26b02e278)\n",
"|---|---|---|\n",
"<center>图8 (左)基于回归的CTPN[3]算法优化anchor (中)基于分割的DB[7]算法优化后处理 (右)回归+分割的SAST[10]算法</center>\n",
"\n",
"<br>\n",
"\n",
"文本检测相关技术将在第二章进行详细解读和实战。\n",
"\n",
"## 2.2 文本识别\n",
"\n",
"文本识别的任务是识别出图像中的文字内容,一般输入来自于文本检测得到的文本框截取出的图像文字区域。文本识别一般可以根据待识别文本形状分为**规则文本识别**和**不规则文本识别**两大类。规则文本主要指印刷字体、扫描文本等,文本大致处在水平线位置;不规则文本往往不在水平位置,存在弯曲、遮挡、模糊等问题。不规则文本场景具有很大的挑战性,也是目前文本识别领域的主要研究方向。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/b292f21e50c94debab7496d4ced96a93774a8525c12346f49cb151bde2a58fe8)\n",
"<center>图9 (左)规则文本 VS. (右)不规则文本</center>\n",
"\n",
"<br>\n",
"\n",
"规则文本识别的算法根据解码方式的不同可以大致分为基于CTC和Sequence2Sequence两种,将网络学习到的序列特征 转化为 最终的识别结果 的处理方式不同。基于CTC的算法以经典的CRNN[11]为代表。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/403ca85c59d344f88d3b1229ca14b1e90c5c73c9f1d248b7aa94103f9d0af597)\n",
"<center>图10 基于CTC的识别算法 VS. 基于Attention的识别算法</center>\n",
"\n",
"不规则文本的识别算法相比更为丰富,如STAR-Net[12]等方法通过加入TPS等矫正模块,将不规则文本矫正为规则的矩形后再进行识别;RARE[13]等基于Attention的方法增强了对序列之间各部分相关性的关注;基于分割的方法将文本行的各字符作为独立个体,相比与对整个文本行做矫正后识别,识别分割出的单个字符更加容易;此外,随着近年来Transfomer[14]的快速发展和在各类任务中的有效性验证,也出现了一批基于Transformer的文本识别算法,这类方法利用transformer结构解决CNN在长依赖建模上的局限性问题,也取得了不错的效果。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/0fa30c3789424473ad9be1c87a4f742c1db69e3defb64651906e5334ed9571a8)\n",
"<center>图11 基于字符分割的识别算法[15]</center>\n",
"\n",
"<br>\n",
"\n",
"文本识别相关技术将在第三章进行详细解读和实战。\n",
"\n",
"## 2.3 文档结构化识别\n",
"\n",
"传统意义上的OCR技术可以解决文字的检测和识别需求,但在实际应用场景中,最终需要获取的往往是结构化的信息,如身份证、发票的信息格式化抽取,表格的结构化识别等等,多在快递单据抽取、合同内容比对、金融保理单信息比对、物流业单据识别等场景下应用。OCR结果+后处理是一种常用的结构化方案,但流程往往比较复杂,并且后处理需要精细设计,泛化性也比较差。在OCR技术逐渐成熟、结构化信息抽取需求日益旺盛的背景下,版面分析、表格识别、关键信息提取等关于智能文档分析的各种技术受到了越来越多的关注和研究。\n",
"\n",
"* **版面分析**\n",
"\n",
"版面分析(Layout Analysis)主要是对文档图像进行内容分类,类别一般可分为纯文本、标题、表格、图片等。现有方法一般将文档中不同的板式当做不同的目标进行检测或分割,如Soto Carlos[16]在目标检测算法Faster R-CNN的基础上,结合上下文信息并利用文档内容的固有位置信息来提高区域检测性能;Sarkar Mausoom[17]等人提出了一种基于先验的分割机制,在非常高的分辨率的图像上训练文档分割模型,解决了过度缩小原始图像导致的密集区域不同结构无法区分进而合并的问题。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/dedb212e8972497998685ff51af7bfe03fdea57f6acd450281ad100807086e1a)\n",
"<center>图12 版面分析任务示意图</center>\n",
"\n",
"<br>\n",
"\n",
"* **表格识别**\n",
"\n",
"表格识别(Table Recognition)的任务就是将文档里的表格信息进行识别和转换到excel文件中。文本图像中表格种类和样式复杂多样,例如不同的行列合并,不同的内容文本类型等,除此之外文档的样式和拍摄时的光照环境等都为表格识别带来了极大的挑战。这些挑战使得表格识别一直是文档理解领域的研究难点。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/47119a2a2f9a45788390d6506f90d5de7449738008aa4c0ab619b18f37bd8d57)\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/22ca5749441441e69dc0eaeb670832a5d0ae0ce522f34731be7d609a2d36e8c1)\n",
"<center>图13 表格识别任务示意图</center>\n",
"\n",
"<br>\n",
"\n",
"表格识别的方法种类较为丰富,早期的基于启发式规则的传统算法,如Kieninger[18]等人提出的T-Rect等算法,一般通过人工设计规则,连通域检测分析处理;近年来随着深度学习的发展,开始涌现一些基于CNN的表格结构识别算法,如Siddiqui Shoaib Ahmed[19]等人提出的DeepTabStR,Raja Sachin[20]等人提出的TabStruct-Net等;此外,随着图神经网络(Graph Neural Network)的兴起,也有一些研究者尝试将图神经网络应用到表格结构识别问题上,基于图神经网络,将表格识别看作图重建问题,如Xue Wenyuan[21]等人提出的TGRNet;基于端到端的方法直接使用网络完成表格结构的HTML表示输出,端到端的方法大多采用Seq2Seq方法来完成表格结构的预测,如一些基于Attention或Transformer的方法,如TableMaster[22]。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/a9a3c91898c84f03b382583859526c4b451ace862dbc4a15838f5dde4d0ea657)\n",
"<center>图14 表格识别方法示意图</center>\n",
"\n",
"<br>\n",
"\n",
"* **关键信息提取**\n",
"\n",
"关键信息提取(Key Information Extraction,KIE)是Document VQA中的一个重要任务,主要从图像中提取所需要的关键信息,如从身份证中提取出姓名和公民身份号码信息,这类信息的种类往往在特定任务下是固定的,但是在不同任务间是不同的。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/8af011647bb4464f80d07f3efeac469baed27c8185ef4c4883a19f40e8ba91f5)\n",
"<center>图15 DocVQA任务示意图</center>\n",
"\n",
"<br>\n",
"\n",
"KIE通常分为两个子任务进行研究:\n",
"\n",
"- SER: 语义实体识别 (Semantic Entity Recognition),对每一个检测到的文本进行分类,如将其分为姓名,身份证。如下图中的黑色框和红色框。\n",
"- RE: 关系抽取 (Relation Extraction),对每一个检测到的文本进行分类,如将其分为问题和的答案。然后对每一个问题找到对应的答案。如下图中的红色框和黑色框分别代表问题和答案,黄色线代表问题和答案之间的对应关系。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/2f1bc1a3e4a341ab9552bbf5f6c2be71ba78d7d65da64818b776efe0691e310b)\n",
"<center>图16 ser与re任务</center>\n",
"\n",
"<br>\n",
"\n",
"一般的KIE方法基于命名实体识别(Named Entity Recognition,NER)[4]来研究,但是这类方法只利用了图像中的文本信息,缺少对视觉和结构信息的使用,因此精度不高。在此基础上,近几年的方法都开始将视觉和结构信息与文本信息融合到一起,按照对多模态信息进行融合时所采用的的原理可以将这些方法分为下面四种:\n",
"\n",
"- 基于Grid的方法\n",
"- 基于Token的方法\n",
"- 基于GCN的方法\n",
"- 基于End to End 的方法\n",
"\n",
"<br>\n",
"\n",
"文档分析相关技术将在第六章进行详细解读和实战。\n",
"\n",
"## 2.4 其他相关技术\n",
"\n",
"前面主要介绍了OCR领域的三种关键技术:文本检测、文本识别、文档结构化识别,更多其他OCR相关前沿技术介绍,包括端到端文本识别、OCR中的图像预处理技术、OCR数据合成等,可参考教程第七章和第八章。\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"# 3. OCR技术的产业实践\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/3d5f18f7598f405884fa2fab041c95ce415af40712e9489996747f9d122c3d90)\n",
"\n",
"> 你是小王,该怎么办? \n",
"> 1. 我不会,我不行,我不干了😭\n",
"> 2. 建议老板找外包公司或者商业化方案,反正花老板的钱😊\n",
"> 3. 网上找找类似项目,面向Github编程😏\n",
"\n",
"<br>\n",
"\n",
"OCR技术最终还是要落到产业实践当中。虽然学术上关于OCR技术的研究很多,OCR技术的商业化应用相比于其他AI技术也已经相对成熟,但在实际的产业应用中,还是存在一些难点与挑战。下文将从技术和产业实践两个角度进行分析。\n",
"\n",
"\n",
"## 3.1 产业实践难点\n",
"\n",
"在实际的产业实践中,开发者常常需要依托开源社区资源启动或推进项目,而开发者使用开源模型又往往面临三大难题:\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/7e5e79240b9c4f13b675b56bc12edf540f159c922bf24e3cbc4a0635a356c7f9)\n",
"<center>图17 OCR技术产业实践三大难题</center>\n",
"\n",
"**1. 找不到、选不出**\n",
"\n",
"开源社区资源丰富,但是信息不对称导致开发者并不能高效地解决痛点问题。一方面,开源社区资源过于丰富,开发者面对一项需求,无法快速从海量的代码仓库中找到匹配业务需求的项目,即存在“找不到”的问题;另一方面,在算法选型时,英文公开数据集上的指标,无法给开发者常常面对的中文场景提供直接的参考,逐个算法验证需要耗费大量时间和人力,且不能保证选出最合适的算法,即“选不出”。\n",
"\n",
"**2. 不适用产业场景**\n",
"\n",
"开源社区中的工作往往更多地偏向效果优化,如学术论文代码开源或复现,一般更侧重算法效果,平衡考虑模型大小和速度的工作相比就少很多,而模型大小和预测耗时在产业实践中是两项不容忽视的指标,其重要程度不亚于模型效果。无论是移动端和服务器端,待识别的图像数目往往非常多,都希望模型更小,精度更高,预测速度更快。GPU太贵,最好使用CPU跑起来更经济。在满足业务需求的前提下,模型越轻量占用的资源越少。\n",
"\n",
"**3. 优化难、训练部署问题多**\n",
"\n",
"直接使用开源算法或模型一般无法直接满足业务需求,实际业务场景中,OCR面临的问题多种多样,业务场景个性化往往需要自定义数据集重新训练,现有的开源项目上,实验各种优化方法的成本较高。此外,OCR应用场景十分丰富,服务端和各种移动端设备上都有着广泛的应用需求,硬件环境多样化就需要支持丰富的部署方式,而开源社区的项目更侧重算法和模型,在预测部署这部分明显支撑不足。要把OCR技术从论文上的算法做到技术落地应用,对开发者的算法和工程能力都有很高的要求。\n",
"\n",
"## 3.2 产业级OCR开发套件PaddleOCR\n",
"\n",
"OCR产业实践需要一套完整全流程的解决方案,来加快研发进度,节约宝贵的研发时间。也就是说,超轻量模型及其全流程解决方案,尤其对于算力、存储空间有限的移动端、嵌入式设备而言,可以说是刚需。\n",
"\n",
"在此背景下,产业级OCR开发套件[PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR)应运而生。\n",
"\n",
"PaddleOCR的建设思路从用户画像和需求出发,依托飞桨核心框架,精选并复现丰富的前沿算法,基于复现的算法研发更适用于产业落地的PP特色模型,并打通训推一体,提供多种预测部署方式,满足实际应用的不同需求场景。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/e09929b4a31e44f9b5e3d542d12411332669d2e1a21d45ad88b1dd91142ec86c)\n",
"<center>图18 PaddleOCR开发套件全景图</center>\n",
"\n",
"<br>\n",
"\n",
"从全景图可以看出,PaddleOCR依托于飞桨核心框架,在模型算法、预训练模型库、工业级部署等层面均提供了丰富的解决方案,并且提供了数据合成、半自动数据标注工具,满足开发者的数据生产需求。\n",
"\n",
"**在模型算法层面**,PaddleOCR对**文字检测识别**和**文档结构化分析**两类任务分别提供了解决方案。在文字检测识别方面,PaddleOCR复现或开源了4种文本检测算法、8种文本识别算法、1种端到端文本识别算法,并在此基础上研发了PP-OCR系列的通用文本检测识别解决方案;在文档结构化分析方面,PaddleOCR提供了版面分析、表格识别、关键信息抽取、命名实体识别等算法,并在此基础提出了PP-Structure文档分析解决方案。丰富的精选算法可以满足开发者不同业务场景的需求,代码框架的统一也方便开发者进行不同算法的优化和性能对比。\n",
"\n",
"**在预训练模型库层面**,基于PP-OCR和PP-Structure解决方案,PaddleOCR研发并开源了适用于产业实践的PP系列特色模型,包括通用、超轻量和多语言的文本检测识别模型,和复杂文档分析模型。PP系列特色模型均在原始算法上进行了深度优化,使其在效果和性能上均能达到产业实用级别,开发者既可以直接应用于业务场景,也可以用业务数据进行简单的finetune,便可以轻松研发出适用于自己业务需求的“实用模型”。\n",
"\n",
"**在工业级部署层面**,PaddleOCR提供了基于Paddle Inference的服务器端预测方案,基于Paddle Serving的服务化部署方案,以及基于Paddle-Lite的端侧部署方案,满足不同硬件环境下的部署需求,同时提供了基于PaddleSlim的模型压缩方案,可以进一步压缩模型大小。以上部署方式都完成了训推一体全流程打通,以保障开发者可以高效部署,稳定可靠。\n",
"\n",
"**在数据工具层面**,PaddleOCR提供了半自动数据标注工具PPOCRLabel和数据合成工具Style-Text,助力开发者更方便的生产模型训练所需的数据集和标注信息。PPOCRLabel作为业界首个开源的半自动OCR数据标注工具,针对标注过程枯燥繁琐、机械性高,大量训练数据所需人工标记,时间金钱成本昂贵的问题,内置PP-OCR模型实现预标注+人工校验的标注模式,可以极大提升标注效率,节省人力成本。数据合成工具Style-Text主要解决实际场景真实数据严重不足,传统合成算法无法合成文字风格(字体、颜色、间距、背景)的问题,只需要少许目标场景图像,就可以批量合成大量与目标场景风格相近的文本图像。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/90a358d6a62c49b7b8db47e18c77878c60f80cf9c81541bfa3befea68d9dbc0f)\n",
"<center>图19 PPOCRLabel使用示意图</center>\n",
"\n",
"<br>\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/b63b10bc530c42bea3d3b923da6000f1cfef006d7eec4ff3bdc0439bd9c333c9)\n",
"<center>图20 Style-Text合成效果示例</center>\n",
"\n",
"<br>\n",
"\n",
"### 3.2.1 PP-OCR与PP-Structrue\n",
"\n",
"PP系列特色模型是飞桨各视觉开发套件针对产业实践需求进行深度优化的模型,力求速度与精度平衡。PaddleOCR中的PP系列特色模型包括针对文字检测识别任务的PP-OCR系列模型和针对文档分析的PP-Structure系列模型。\n",
"\n",
"**(1)PP-OCR中英文模型**\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/3372558042044d43983b815069e1e43cb84432b993ed400f946976e75bd51f38)\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/f0a0b936382c42dd8809e98759b4c84434d79386606b4d5b8a86416db6dbaeee)\n",
"<center>图21 PP-OCR中英文模型识别结果示例</center>\n",
"\n",
"<br>\n",
"\n",
"PP-OCR中英文模型采用的典型的两阶段OCR算法,即检测模型+识别模型的组成方式,具体的算法框架如下:\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/8af1371b5e3c486bb90a041903200c7c666c8bbc98c245dc802ff8c4da98617e)\n",
"<center>图22 PP-OCR系统pipeline示意图</center>\n",
"\n",
"<br>\n",
"\n",
"可以看到,除输入输出外,PP-OCR核心框架包含了3个模块,分别是:文本检测模块、检测框矫正模块、文本识别模块。\n",
"- 文本检测模块:核心是一个基于[DB](https://arxiv.org/abs/1911.08947)检测算法训练的文本检测模型,检测出图像中的文字区域;\n",
"- 检测框矫正模块:将检测到的文本框输入检测框矫正模块,在这一阶段,将四点表示的文本框矫正为矩形框,方便后续进行文本识别,另一方面会进行文本方向判断和校正,例如如果判断文本行是倒立的情况,则会进行转正,该功能通过训练一个文本方向分类器实现;\n",
"- 文本识别模块:最后文本识别模块对矫正后的检测框进行文本识别,得到每个文本框内的文字内容,PP-OCR中使用的经典文本识别算法[CRNN](https://arxiv.org/abs/1507.05717)。\n",
"\n",
"PaddleOCR先后推出了PP-OCR[23]和PP-OCRv2[24]模型。\n",
"\n",
"PP-OCR模型分为mobile版(轻量版)和server版(通用版),其中mobile版模型主要基于轻量级骨干网络MobileNetV3进行优化,优化后模型(检测模型+文本方向分类模型+识别模型)大小仅8.1M,CPU上平均单张图像预测耗时350ms,T4 GPU上约110ms,裁剪量化后,可在精度不变的情况下进一步压缩到3.5M,便于端侧部署,在骁龙855上测试预测耗时仅260ms。更多的PP-OCR评估数据可参考[benchmark](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.2/doc/doc_ch/benchmark.md)。\n",
"\n",
"PP-OCRv2保持了PP-OCR的整体框架,主要做了效果上的进一步策略优化。提升包括3个方面:\n",
"- 在模型效果上,相对于PP-OCR mobile版本提升超7%;\n",
"- 在速度上,相对于PP-OCR server版本提升超过220%;\n",
"- 在模型大小上,11.6M的总大小,服务器端和移动端都可以轻松部署。\n",
"\n",
"PP-OCR和PP-OCRv2的具体优化策略将在第四章中进行详细解读。\n",
"\n",
"除了中英文模型,PaddleOCR也基于不同的数据集训练并开源了英文数字模型、多语言识别模型,以上均为超轻量模型,适用于不同的语言场景。\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/5978652a826647b98344cf61aa1c2027662af989b73e4a0e917d83718422eeb0)\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/1a8a8e24b5a440d388dae767adf0ea9c049335b04e964abbb176f58c5b028d7e)\n",
"<center>图23 PP-OCR的英文数字模型和多语言模型识别效果示意图</center>\n",
"\n",
"<br>\n",
"\n",
"**(2)PP-Structure文档分析模型**\n",
"\n",
"PP-Structure支持版面分析(layout analysis)、表格识别(table recognition)、文档视觉问答(DocVQA)三种子任务。\n",
"\n",
"PP-Structure核心功能点如下:\n",
"- 支持对图片形式的文档进行版面分析,可以划分文字、标题、表格、图片以及列表5类区域(与Layout-Parser联合使用)\n",
"- 支持文字、标题、图片以及列表区域提取为文字字段(与PP-OCR联合使用)\n",
"- 支持表格区域进行结构化分析,最终结果输出Excel文件\n",
"- 支持Python whl包和命令行两种方式,简单易用\n",
"- 支持版面分析和表格结构化两类任务自定义训练\n",
"- 支持VQA任务-SER和RE\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/129708c265644dbc90d6c8f7db224b3a6f11f37bb586463a82e7ccb50bcc2e76)\n",
"<center>图24 PP-Structure系统示意图(本图仅含版面分析+表格识别)</center>\n",
"\n",
"<br>\n",
"\n",
"PP-Structure的具体方案将在第六章中进行详细解读。\n",
"\n",
"### 3.2.2 工业级部署方案\n",
"\n",
"飞桨支持全流程、全场景推理部署,模型来源主要分为三种,第一种使用PaddlePaddle API构建网络结构进行训练所得,第二种是基于飞桨套件系列,飞桨套件提供了丰富的模型库、简洁易用的API,具备开箱即用,包括视觉模型库PaddleCV、智能语音库PaddleSpeech以及自然语言处理库PaddleNLP等,第三种采用X2Paddle工具从第三方框架(PyTorh、ONNX、TensorFlow等)产出的模型。\n",
"\n",
"飞桨模型可以选用PaddleSlim工具进行压缩、量化以及蒸馏,支持五种部署方案,分别为服务化Paddle Serving、服务端/云端Paddle Inference、移动端/边缘端Paddle Lite、网页前端Paddle.js, 对于Paddle不支持的硬件,比如MCU、地平线、鲲云等国产芯片,可以借助Paddle2ONNX转化为支持ONNX的第三方框架。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/c9ffe78e7db14e4eb103e7f393a16fbf2ab438540250474a8e0e7adc4aeb7ee0)\n",
"<center>图25 飞桨支持部署方式</center>\n",
"\n",
"<br>\n",
"\n",
"Paddle Inference支持服务端和云端部署,具备高性能与通用性,针对不同平台和不同应用场景进行了深度的适配和优化,Paddle Inference是飞桨的原生推理库,保证模型在服务器端即训即用,快速部署,适用于高性能硬件上使用多种应用语言环境部署算法复杂的模型,硬件覆盖x86 CPU、Nvidia GPU、以及百度昆仑XPU、华为昇腾等AI加速器。\n",
"\n",
"Paddle Lite 是端侧推理引擎,具有轻量化和高性能特点,针对端侧设备和各应用场景进行了深度的设配和优化。当前支持Android、IOS、嵌入式Linux设备、macOS 等多个平台,硬件覆盖ARM CPU和GPU、X86 CPU和新硬件如百度昆仑、华为昇腾与麒麟、瑞芯微等。\n",
"\n",
"Paddle Serving是一套高性能服务框架,旨在帮助用户几个步骤快速将模型在云端服务化部署。目前Paddle Serving支持自定义前后处理、模型组合、模型热加载更新、多机多卡多模型、分布式推理、K8S部署、安全网关和模型加密部署、支持多语言多客户端访问等功能,Paddle Serving官方还提供了包括PaddleOCR在内的40多种模型的部署示例,以帮助用户更快上手。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/4d8063d74194434ea9b7c9f81c7fbdfd2131e13770124d2e99c1b9670f12e019)\n",
"<center>图26 飞桨支持部署方式</center>\n",
"\n",
"<br>\n",
"\n",
"以上部署方案将在第五章中基于PP-OCRv2模型进行详细解读与实战。"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"# 4. 总结\n",
"\n",
"本节首先介绍了OCR技术的应用场景和前沿算法,然后分析了OCR技术在产业实践中的难点与三大挑战。\n",
"\n",
"本教程后续章节内容安排如下:\n",
"\n",
"* 第二、三章分别介绍检测、识别技术并实践;\n",
"* 第四章介绍PP-OCR优化策略; \n",
"* 第五章进行预测部署实战; \n",
"* 第六章介绍文档结构化; \n",
"* 第七章介绍端到端、数据预处理、数据合成等其他OCR相关算法; \n",
"* 第八章介绍OCR相关数据集和数据合成工具。\n",
"\n",
"# 参考文献\n",
"\n",
"[1] Liao, Minghui, et al. \"Textboxes: A fast text detector with a single deep neural network.\" Thirty-first AAAI conference on artificial intelligence. 2017.\n",
"\n",
"[2] Liu W, Anguelov D, Erhan D, et al. Ssd: Single shot multibox detector[C]//European conference on computer vision. Springer, Cham, 2016: 21-37.\n",
"\n",
"[3] Tian, Zhi, et al. \"Detecting text in natural image with connectionist text proposal network.\" European conference on computer vision. Springer, Cham, 2016.\n",
"\n",
"[4] Ren S, He K, Girshick R, et al. Faster r-cnn: Towards real-time object detection with region proposal networks[J]. Advances in neural information processing systems, 2015, 28: 91-99.\n",
"\n",
"[5] Zhou, Xinyu, et al. \"East: an efficient and accurate scene text detector.\" Proceedings of the IEEE conference on Computer Vision and Pattern Recognition. 2017.\n",
"\n",
"[6] Wang, Wenhai, et al. \"Shape robust text detection with progressive scale expansion network.\" Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2019.\n",
"\n",
"[7] Liao, Minghui, et al. \"Real-time scene text detection with differentiable binarization.\" Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 34. No. 07. 2020.\n",
"\n",
"[8] Deng, Dan, et al. \"Pixellink: Detecting scene text via instance segmentation.\" Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 32. No. 1. 2018.\n",
"\n",
"[9] He K, Gkioxari G, Dollár P, et al. Mask r-cnn[C]//Proceedings of the IEEE international conference on computer vision. 2017: 2961-2969.\n",
"\n",
"[10] Wang P, Zhang C, Qi F, et al. A single-shot arbitrarily-shaped text detector based on context attended multi-task \n",
"learning[C]//Proceedings of the 27th ACM international conference on multimedia. 2019: 1277-1285.\n",
"\n",
"[11] Shi, B., Bai, X., & Yao, C. (2016). An end-to-end trainable neural network for image-based sequence recognition and its application to scene text recognition. IEEE transactions on pattern analysis and machine intelligence, 39(11), 2298-2304.\n",
"\n",
"[12] Star-Net Max Jaderberg, Karen Simonyan, Andrew Zisserman, et al. Spa- tial transformer networks. In Advances in neural information processing systems, pages 2017–2025, 2015.\n",
"\n",
"[13] Shi, B., Wang, X., Lyu, P., Yao, C., & Bai, X. (2016). Robust scene text recognition with automatic rectification. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4168-4176).\n",
"\n",
"[14] Sheng, F., Chen, Z., & Xu, B. (2019, September). NRTR: A no-recurrence sequence-to-sequence model for scene text recognition. In 2019 International Conference on Document Analysis and Recognition (ICDAR) (pp. 781-786). IEEE.\n",
"\n",
"[15] Lyu P, Liao M, Yao C, et al. Mask textspotter: An end-to-end trainable neural network for spotting text with arbitrary shapes[C]//Proceedings of the European Conference on Computer Vision (ECCV). 2018: 67-83.\n",
"\n",
"[16] Soto C, Yoo S. Visual detection with context for document layout analysis[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019: 3464-3470.\n",
"\n",
"[17] Sarkar M, Aggarwal M, Jain A, et al. Document Structure Extraction using Prior based High Resolution Hierarchical Semantic Segmentation[C]//European Conference on Computer Vision. Springer, Cham, 2020: 649-666.\n",
"\n",
"[18] Kieninger T, Dengel A. A paper-to-HTML table converting system[C]//Proceedings of document analysis systems (DAS). 1998, 98: 356-365.\n",
"\n",
"[19] Siddiqui S A, Fateh I A, Rizvi S T R, et al. Deeptabstr: Deep learning based table structure recognition[C]//2019 International Conference on Document Analysis and Recognition (ICDAR). IEEE, 2019: 1403-1409.\n",
"\n",
"[20] Raja S, Mondal A, Jawahar C V. Table structure recognition using top-down and bottom-up cues[C]//European Conference on Computer Vision. Springer, Cham, 2020: 70-86.\n",
"\n",
"[21] Xue W, Yu B, Wang W, et al. TGRNet: A Table Graph Reconstruction Network for Table Structure Recognition[J]. arXiv preprint arXiv:2106.10598, 2021.\n",
"\n",
"[22] Ye J, Qi X, He Y, et al. PingAn-VCGroup's Solution for ICDAR 2021 Competition on Scientific Literature Parsing Task B: Table Recognition to HTML[J]. arXiv preprint arXiv:2105.01848, 2021.\n",
"\n",
"[23] Du Y, Li C, Guo R, et al. PP-OCR: A practical ultra lightweight OCR system[J]. arXiv preprint arXiv:2009.09941, 2020.\n",
"\n",
"[24] Du Y, Li C, Guo R, et al. PP-OCRv2: Bag of Tricks for Ultra Lightweight OCR System[J]. arXiv preprint arXiv:2109.03144, 2021.\n",
"\n"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "py35-paddle1.2.0"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.4"
}
},
"nbformat": 4,
"nbformat_minor": 1
}
{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"# 文本检测FAQ\n",
"\n",
"本节罗列一些开发者们使用PaddleOCR的文本检测模型常遇到的一些问题,并给出相应的问题解决方法或建议。\n",
"\n",
"FAQ分两个部分来介绍,分别是:\n",
" - 文本检测训练相关\n",
" - 文本检测预测相关"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 1. 文本检测训练相关FAQ\n",
"\n",
"**1.1 PaddleOCR提供的文本检测算法包括哪些?**\n",
"\n",
"**A**:PaddleOCR中包含多种文本检测模型,包括基于回归的文本检测方法EAST、SAST,和基于分割的文本检测方法DB,PSENet。\n",
"\n",
"\n",
"**1.2:请问PaddleOCR项目中的中文超轻量和通用模型用了哪些数据集?训练多少样本,gpu什么配置,跑了多少个epoch,大概跑了多久?**\n",
"\n",
"**A**:对于超轻量DB检测模型,训练数据包括开源数据集lsvt,rctw,CASIA,CCPD,MSRA,MLT,BornDigit,iflytek,SROIE和合成的数据集等,总数据量越10W,数据集分为5个部分,训练时采用随机采样策略,在4卡V100GPU上约训练500epoch,耗时3天。\n",
"\n",
"\n",
"**1.3 文本检测训练标签是否需要具体文本标注,标签中的”###”是什么意思?**\n",
"\n",
"**A**:文本检测训练只需要文本区域的坐标即可,标注可以是四点或者十四点,按照左上,右上,右下,左下的顺序排列。PaddleOCR提供的标签文件中包含文本字段,对于文本区域文字不清晰会使用###代替。训练检测模型时,不会用到标签中的文本字段。\n",
" \n",
"**1.4 对于文本行较紧密的情况下训练的文本检测模型效果较差?**\n",
"\n",
"**A**:使用基于分割的方法,如DB,检测密集文本行时,最好收集一批数据进行训练,并且在训练时,并将生成二值图像的[shrink_ratio](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/ppocr/data/imaug/make_shrink_map.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L37)参数调小一些。另外,在预测的时候,可以适当减小[unclip_ratio](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L59)参数,unclip_ratio参数值越大检测框就越大。\n",
"\n",
"\n",
"**1.5 对于一些尺寸较大的文档类图片, DB在检测时会有较多的漏检,怎么避免这种漏检的问题呢?**\n",
"\n",
"**A**:首先,需要确定是模型没有训练好的问题还是预测时处理的问题。如果是模型没有训练好,建议多加一些数据进行训练,或者在训练的时候多加一些数据增强。\n",
"如果是预测图像过大的问题,可以增大预测时输入的最长边设置参数[det_limit_side_len](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/tools/infer/utility.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L47),默认为960。\n",
"其次,可以通过可视化后处理的分割图观察漏检的文字是否有分割结果,如果没有分割结果,说明是模型没有训练好。如果有完整的分割区域,说明是预测后处理的问题,建议调整[DB后处理参数](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/tools/infer/utility.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L51-L53)。\n",
"\n",
"\n",
"**1.6 DB模型弯曲文本(如略微形变的文档图像)漏检问题?**\n",
"\n",
"**A**: DB后处理中计算文本框平均得分时,是求rectangle区域的平均分数,容易造成弯曲文本漏检,已新增求polygon区域的平均分数,会更准确,但速度有所降低,可按需选择,在相关pr中可查看[可视化对比效果](https://github.com/PaddlePaddle/PaddleOCR/pull/2604)。该功能通过参数 [det_db_score_mode](https://github.com/PaddlePaddle/PaddleOCR/blob/release/2.1/tools/infer/utility.py#L51)进行选择,参数值可选[`fast`(默认)、`slow`],`fast`对应原始的rectangle方式,`slow`对应polygon方式。感谢用户[buptlihang](https://github.com/buptlihang)提[pr](https://github.com/PaddlePaddle/PaddleOCR/pull/2574)帮助解决该问题。\n",
"\n",
"\n",
"**1.7 简单的对于精度要求不高的OCR任务,数据集需要准备多少张呢?**\n",
"\n",
"**A**:(1)训练数据的数量和需要解决问题的复杂度有关系。难度越大,精度要求越高,则数据集需求越大,而且一般情况实际中的训练数据越多效果越好。\n",
"\n",
"(2)对于精度要求不高的场景,检测任务和识别任务需要的数据量是不一样的。对于检测任务,500张图像可以保证基本的检测效果。对于识别任务,需要保证识别字典中每个字符出现在不同场景的行文本图像数目需要大于200张(举例,如果有字典中有5个字,每个字都需要出现在200张图片以上,那么最少要求的图像数量应该在200-1000张之间),这样可以保证基本的识别效果。\n",
"\n",
"\n",
"**1.8 当训练数据量少时,如何获取更多的数据?**\n",
"\n",
"**A**:当训练数据量少时,可以尝试以下三种方式获取更多的数据:(1)人工采集更多的训练数据,最直接也是最有效的方式。(2)基于PIL和opencv基本图像处理或者变换。例如PIL中ImageFont, Image, ImageDraw三个模块将文字写到背景中,opencv的旋转仿射变换,高斯滤波等。(3)利用数据生成算法合成数据,例如pix2pix等算法。\n",
"\n",
"\n",
"**1.9 如何更换文本检测/识别的backbone?**\n",
"\n",
"A:无论是文字检测,还是文字识别,骨干网络的选择是预测效果和预测效率的权衡。一般,选择更大规模的骨干网络,例如ResNet101_vd,则检测或识别更准确,但预测耗时相应也会增加。而选择更小规模的骨干网络,例如MobileNetV3_small_x0_35,则预测更快,但检测或识别的准确率会大打折扣。幸运的是不同骨干网络的检测或识别效果与在ImageNet数据集图像1000分类任务效果正相关。飞桨图像分类套件PaddleClas汇总了ResNet_vd、Res2Net、HRNet、MobileNetV3、GhostNet等23种系列的分类网络结构,在上述图像分类任务的top1识别准确率,GPU(V100和T4)和CPU(骁龙855)的预测耗时以及相应的117个预训练模型下载地址。\n",
"\n",
"(1)文字检测骨干网络的替换,主要是确定类似与ResNet的4个stages,以方便集成后续的类似FPN的检测头。此外,对于文字检测问题,使用ImageNet训练的分类预训练模型,可以加速收敛和效果提升。\n",
"\n",
"(2)文字识别的骨干网络的替换,需要注意网络宽高stride的下降位置。由于文本识别一般宽高比例很大,因此高度下降频率少一些,宽度下降频率多一些。可以参考[PaddleOCR中MobileNetV3骨干网络的改动](https://github.com/PaddlePaddle/PaddleOCR/blob/release%2F2.3/ppocr/modeling/backbones/rec_mobilenet_v3.py)。\n",
"\n",
"\n",
"**1.10 如何对检测模型finetune,比如冻结前面的层或某些层使用小的学习率学习?**\n",
"\n",
"**A**:如果是冻结某些层,可以将变量的stop_gradient属性设置为True,这样计算这个变量之前的所有参数都不会更新了,参考:https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/faq/train_cn.html#id4\n",
"\n",
"如果对某些层使用更小的学习率学习,静态图里还不是很方便,一个方法是在参数初始化的时候,给权重的属性设置固定的学习率,参考:https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api/paddle/fluid/param_attr/ParamAttr_cn.html#paramattr\n",
"\n",
"实际上我们实验发现,直接加载模型去fine-tune,不设置某些层不同学习率,效果也都不错\n",
"\n",
"**1.11 DB的预处理部分,图片的长和宽为什么要处理成32的倍数?**\n",
"\n",
"**A**:和网络下采样的倍数(stride)有关。以检测中的resnet骨干网络为例,图像输入网络之后,需要经过5次2倍降采样,共32倍,因此建议输入的图像尺寸为32的倍数。\n",
"\n",
"\n",
"**1.12 在PP-OCR系列的模型中,文本检测的骨干网络为什么没有使用SEBlock?**\n",
"\n",
"**A**:SE模块是MobileNetV3网络一个重要模块,目的是估计特征图每个特征通道重要性,给特征图每个特征分配权重,提高网络的表达能力。但是,对于文本检测,输入网络的分辨率比较大,一般是640\\*640,利用SE模块估计特征图每个特征通道重要性比较困难,网络提升能力有限,但是该模块又比较耗时,因此在PP-OCR系统中,文本检测的骨干网络没有使用SE模块。实验也表明,当去掉SE模块,超轻量模型大小可以减小40%,文本检测效果基本不受影响。详细可以参考PP-OCR技术文章,https://arxiv.org/abs/2009.09941.\n",
"\n",
"\n",
"**1.13 PP-OCR检测效果不好,该如何优化?**\n",
"\n",
"A: 具体问题具体分析:\n",
"- 如果在你的场景上检测效果不可用,首选是在你的数据上做finetune训练;\n",
"- 如果图像过大,文字过于密集,建议不要过度压缩图像,可以尝试修改检测预处理的resize逻辑,防止图像被过度压缩;\n",
"- 检测框大小过于紧贴文字或检测框过大,可以调整db_unclip_ratio这个参数,加大参数可以扩大检测框,减小参数可以减小检测框大小;\n",
"- 检测框存在很多漏检问题,可以减小DB检测后处理的阈值参数det_db_box_thresh,防止一些检测框被过滤掉,也可以尝试设置det_db_score_mode为'slow';\n",
"- 其他方法可以选择use_dilation为True,对检测输出的feature map做膨胀处理,一般情况下,会有效果改善;\n",
"\n",
"\n",
"## 2. 文本检测预测相关FAQ\n",
"\n",
"**2.1 DB有些框太贴文本了反而去掉了一些文本的边角影响识别,这个问题有什么办法可以缓解吗?**\n",
"\n",
"**A**:可以把后处理的参数[unclip_ratio](https://github.com/PaddlePaddle/PaddleOCR/blob/d80afce9b51f09fd3d90e539c40eba8eb5e50dd6/tools/infer/utility.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L52)适当调大一点,该参数越大文本框越大。\n",
"\n",
"\n",
"**2.2 为什么PaddleOCR检测预测是只支持一张图片测试?即test_batch_size_per_card=1**\n",
"\n",
"**A**:预测的时候,对图像等比例缩放,最长边960,不同图像等比例缩放后长宽不一致,无法组成batch,所以设置为test_batch_size为1。\n",
"\n",
"\n",
"**2.3 在CPU上加速PaddleOCR的文本检测模型预测?**\n",
"\n",
"**A**:x86 CPU可以使用mkldnn(OneDNN)进行加速;在支持mkldnn加速的CPU上开启[enable_mkldnn](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/tools/infer/utility.py#L105)参数。另外,配合增加CPU上预测使用的[线程数num_threads](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/tools/infer/utility.py#L106),可以有效加快CPU上的预测速度。\n",
"\n",
"**2.4 在GPU上加速PaddleOCR的文本检测模型预测?**\n",
"\n",
"**A**:GPU加速预测推荐使用TensorRT。\n",
"- 1. 从[链接](https://paddleinference.paddlepaddle.org.cn/master/user_guides/download_lib.html)下载带TensorRT的Paddle安装包或者预测库。\n",
"- 2. 从Nvidia官网下载TensorRT版本,注意下载的TensorRT版本与paddle安装包中编译的TensorRT版本一致。\n",
"- 3. 设置环境变量LD_LIBRARY_PATH,指向TensorRT的lib文件夹\n",
"```\n",
"export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<TensorRT-${version}/lib>\n",
"```\n",
"- 4. 开启PaddleOCR预测的[tensorrt选项](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/tools/infer/utility.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L38)。\n",
"\n",
"**2.5 如何在移动端部署PaddleOCR模型?**\n",
"\n",
"**A**: 飞桨Paddle有专门针对移动端部署的工具[PaddleLite](https://github.com/PaddlePaddle/Paddle-Lite),并且PaddleOCR提供了DB+CRNN为demo的android arm部署代码,参考[链接](https://github.com/PaddlePaddle/PaddleOCR/blob/release%2F2.3/deploy/lite/readme.md)。\n",
"\n",
"\n",
"**2.6 如何使用PaddleOCR多进程预测?**\n",
"\n",
"**A**: 近期PaddleOCR新增了[多进程预测控制参数](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/tools/infer/utility.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L111),`use_mp`表示是否使用多进程,`total_process_num`表示在使用多进程时的进程数。具体使用方式请参考[文档](https://github.com/PaddlePaddle/PaddleOCR/blob/release%2F2.3/doc/doc_ch/inference.md#1-%E8%B6%85%E8%BD%BB%E9%87%8F%E4%B8%AD%E6%96%87ocr%E6%A8%A1%E5%9E%8B%E6%8E%A8%E7%90%86)。\n",
"\n",
"**2.7 预测时显存爆炸、内存泄漏问题?**\n",
"\n",
"**A**: 如果是训练模型的预测,由于模型太大或者输入图像太大导致显存不够用,可以参考代码在主函数运行前加上paddle.no_grad(),即可减小显存占用。如果是inference模型预测时显存占用过高,可以配置Config时,加入[config.enable_memory_optim()](https://github.com/PaddlePaddle/PaddleOCR/blob/8b656a3e13631dfb1ac21d2095d4d4a4993ef710/tools/infer/utility.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L267)用于减小内存占用。\n",
"\n",
"另外关于使用Paddle预测时出现内存泄漏的问题,建议安装paddle最新版本,内存泄漏已修复。"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "py35-paddle1.2.0"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.4"
}
},
"nbformat": 4,
"nbformat_minor": 1
}
因为 它太大了无法显示 source diff 。你可以改为 查看blob
{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"# OCR七日课之文本检测综述\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 1. 文本检测\n",
"\n",
"文本检测任务是找出图像或视频中的文字位置。不同于目标检测任务,目标检测不仅要解决定位问题,还要解决目标分类问题。\n",
"\n",
"文本在图像中的表现形式可以视为一种‘目标‘,通用的目标检测的方法也适用于文本检测,从任务本身上来看:\n",
"\n",
"- 目标检测:给定图像或者视频,找出目标的位置(box),并给出目标的类别;\n",
"- 文本检测:给定输入图像或者视频,找出文本的区域,可以是单字符位置或者整个文本行位置;\n",
"\n",
"\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/af2d8eca913a4d5a968945ae6cac180b009c6cc94abc43bfbaf1ba6a3de98125\" width=\"400\" ></center>\n",
"\n",
"<br><center>图1 目标检测示意图</center>\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/400b9100573b4286b40b0a668358bcab9627f169ab934133a1280361505ddd33\" width=\"1000\" ></center>\n",
"\n",
"<br><center>图2 文本检测示意图</center>\n",
"\n",
"目标检测和文本检测同属于“定位”问题。但是文本检测无需对目标分类,并且文本形状复杂多样。\n",
"\n",
"当前所说的文本检测一般是自然场景文本检测,其难点在于:\n",
"\n",
"1. 自然场景中文本具有多样性:文本检测受到文字颜色、大小、字体、形状、方向、语言、以及文本长度的影响;\n",
"2. 复杂的背景和干扰;文本检测受到图像失真,模糊,低分辨率,阴影,亮度等因素的影响;\n",
"3. 文本密集甚至重叠会影响文字的检测;\n",
"4. 文字存在局部一致性,文本行的一小部分,也可视为是独立的文本;\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/072f208f2aff47e886cf2cf1378e23c648356686cf1349c799b42f662d8ced00\"\n",
"width=\"1000\" ></center>\n",
"\n",
"<br><center>图3 文本检测场景</center>\n",
"\n",
"针对以上问题,衍生了很多基于深度学习的文本检测算法,解决自然场景文字检测问题,这些方法可以分为基于回归和基于分割的文本检测方法。\n",
"\n",
"下一节将简要介绍基于深度学习技术的经典文字检测算法。"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 2. 文本检测方法介绍\n",
"\n",
"\n",
"近些年来基于深度学习的文本检测算法层出不穷,这些方法大致可以分为两类:\n",
"1. 基于回归的文本检测方法\n",
"2. 基于分割的文本检测方法\n",
"\n",
"\n",
"本节筛选了2017-2021年的常用文本检测方法,按照如上两类方法分类如下表格所示:\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/22314238b70b486f942701107ffddca48b87235a473c4d8db05b317f132daea0\"\n",
"width=\"600\" ></center>\n",
"<br><center>图4 文本检测算法</center>\n",
"\n",
"\n",
"### 2.1 基于回归的文本检测\n",
"\n",
"基于回归文本检测方法和目标检测算法的方法相似,文本检测方法只有两个类别,图像中的文本视为待检测的目标,其余部分视为背景。\n",
"\n",
"#### 2.1.1 水平文本检测\n",
"\n",
"早期基于深度学习的文本检测算法是从目标检测的方法改进而来,支持水平文本检测。比如TextBoxes算法基于SSD算法改进而来,CTPN根据二阶段目标检测Fast-RCNN算法改进而来。\n",
"\n",
"在TextBoxes[1]算法根据一阶段目标检测器SSD调整,将默认文本框更改为适应文本方向和宽高比的规格的四边形,提供了一种端对端训练的文字检测方法,并且无需复杂的后处理。\n",
"- 采用更大长宽比的预选框\n",
"- 卷积核从3x3变成了1x5,更适合长文本检测\n",
"- 采用多尺度输入\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/3864ccf9d009467cbc04225daef0eb562ac0c8c36f9b4f5eab036c319e5f05e7\" width=\"1000\" ></center>\n",
"<br><center>图5 textbox框架图</center>\n",
"\n",
"CTPN[3]基于Fast-RCNN算法,扩展RPN模块并且设计了基于CRNN的模块让整个网络从卷积特征中检测到文本序列,二阶段的方法通过ROI Pooling获得了更准确的特征定位。但是TextBoxes和CTPN只支持检测横向文本。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/452833c2016e4cf7b35291efd09740c13c4bfb8f7c56446b8f7a02fc7eb3e901\" width=\"1000\" ></center>\n",
"<br><center>图6 CTPN框架图</center>\n",
"\n",
"#### 2.1.2 任意角度文本检测\n",
"\n",
"TextBoxes++[2]在TextBoxes基础上进行改进,支持检测任意角度的文本。从结构上来说,不同于TextBoxes,TextBoxes++针对多角度文本进行检测,首先修改预选框的宽高比,调整宽高比aspect ratio为1、2、3、5、1/2、1/3、1/5。其次是将$1*5$的卷积核改为$3*5$,更好的学习倾斜文本的特征;最后,TextBoxes++的输出旋转框的表示信息。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/ae96e3acbac04be296b6d54a4d72e5881d592fcc91f44882b24bc7d38b9d2658\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图7 TextBoxes++框架图</center>\n",
"\n",
"\n",
"EAST[4]针对倾斜文本的定位问题,提出了two-stage的文本检测方法,包含 FCN特征提取和NMS部分。EAST提出了一种新的文本检测pipline结构,可以端对端训练并且支持检测任意朝向的文本,并且具有结构简单,性能高的特点。FCN支持输出倾斜的矩形框和水平框,可以自由选择输出格式。\n",
"- 如果输出检测形状为RBox,则输出Box旋转角度以及AABB文本形状信息,AABB表示到文本框上下左右边的偏移。RBox可以旋转矩形的文本。\n",
"- 如果输出检测框为四点框,则输出的最后一个维度为8个数字,表示从四边形的四个角顶点的位置偏移。该输出方式可以预测不规则四边形的文本。\n",
"\n",
"考虑到FCN输出的文本框是比较冗余的,比如一个文本区域的邻近的像素生成的框重合度较高,但不是同一个文本生成的检测框,重合度都很小,因此EAST提出先按行合并预测框,最后再把剩下的四边形用原始的NMS筛选。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/d7411ada08714adab73fa0edf7555a679327b71e29184446a33d81cdd910e4fc\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图8 EAST框架图</center> \n",
"\n",
"\n",
"MOST[15]提出TFAM模块动态的调整粗粒度的检测结果的感受野,另外提出PA-NMS根据位置信息合并可靠的检测预测结果。此外,训练中还提出 Instance-wise IoU 损失函数,用于平衡训练,以处理不同尺度的文本实例。该方法可以和EAST方法结合,在检测极端长宽比和不同尺度的文本有更好的检测效果和性能。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/73052d9439714bba86ffe4a959d58c523b07baf3f1d74882b4517e71f5a645fe\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图9 MOST框架图</center>\n",
"\n",
"\n",
"#### 2.1.3 弯曲文本检测\n",
"\n",
"利用回归的方法解决弯曲文本的检测问题,一个简单的思路是用多点坐标描述弯曲文本的边界多边形,然后直接预测多边形的顶点坐标。\n",
"\n",
"CTD[6]提出了直接预测弯曲文本14个顶点的边界多边形,网络中利用Bi-LSTM[13]层以细化顶点的预测坐标,实现了基于回归方法的弯曲文本检测。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/6e33d76ebb814cac9ebb2942b779054af160857125294cd69481680aca2fa98a\"\n",
"width=\"600\" ></center>\n",
"<br><center>图10 CTD框架图</center>\n",
"\n",
"\n",
"\n",
"LOMO[19]针对长文本和弯曲文本问题,提出迭代的优化文本定位特征获取更精细的文本定位,该方法包括三个部分,坐标回归模块DR,迭代优化模块IRM以及任意形状表达模块SEM。分别用于生成文本大致区域,迭代优化文本定位特征,预测文本区域、文本中心线以及文本边界。迭代的优化文本特征可以更好的解决长文本定位问题以及获得更精确的文本区域定位。\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/e90adf3ca25a45a0af0b84a181fbe2c4954be1fcca8f4049957128548b7131ef\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图11 LOMO框架图</center>\n",
"\n",
"\n",
"Contournet[18]基于提出对文本轮廓点建模获取弯曲文本检测框,该方法首先使用Adaptive-RPN获取文本区域的proposal特征,然后设计了局部正交纹理感知LOTM模块学习水平与竖直方向的纹理特征,并用轮廓点表示,最后,通过同时考虑两个正交方向上的特征响应,利用Point Re-Scoring算法可以有效地滤除强单向或弱正交激活的预测,最终文本轮廓可以用一组高质量的轮廓点表示出来。\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/1f59ab5db899412f8c70ba71e8dd31d4ea9480d6511f498ea492c97dd2152384\"\n",
"width=\"600\" ></center>\n",
"<br><center>图12 Contournet框架图</center>\n",
"\n",
"\n",
"PCR[14]提出渐进式的坐标回归处理弯曲文本检测问题,总体分为三个阶段,首先大致检测到文本区域,获得文本框,另外通过所设计的Contour Localization Mechanism预测文本最小包围框的角点坐标,然后通过叠加多个CLM模块和RCLM模块预测得到弯曲文本。该方法利用文本轮廓信息聚合得到丰富的文本轮廓特征表示,不仅能抑制冗余的噪声点对坐标回归的影响,还能更精确的定位文本区域。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/c677c4602cee44999ae4b38bd780b69795887f2ae10747968bb084db6209b6cc\"\n",
"width=\"600\" ></center>\n",
"<br><center>图13 PCR框架图</center>\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"\n",
"### 2.2 基于分割的文本检测\n",
"\n",
"基于回归的方法虽然在文本检测上取得了很好的效果,但是对解决弯曲文本往往难以得到平滑的文本包围曲线,并且模型较为复杂不具备性能优势。于是研究者们提出了基于图像分割的文本分割方法,先从像素层面做分类,判别每一个像素点是否属于一个文本目标,得到文本区域的概率图,通过后处理方式得到文本分割区域的包围曲线。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/fb9e50c410984c339481869ba11c1f39f80a4d74920b44b084601f2f8a23099f\"\n",
"width=\"600\" ></center>\n",
"<br><center>图14 文本分割算法示意图</center>\n",
"\n",
"\n",
"此类方法通常是基于分割的方法实现文本检测,基于分割的方法对不规则形状的文本检测有着天然的优势。基于分割的文本检测方法主体思想为,通过分割方法得到图像中文本区域,再利用opencv,polygon等后处理得到文本区域的最小包围曲线。\n",
"\n",
"\n",
"Pixellink[7]采用分割的方法解决文本检测问题,分割对象为文本区域,将同属于一个文本行(单词)中的像素链接在一起来分割文本,直接从分割结果中提取文本边界框,无需位置回归就能达到基于回归的文本检测的效果。但是基于分割的方法存在一个问题,对于位置相近的文本,文本分割区域容易出现“粘连“问题。Wu, Yue等人[8]提出分割文本的同时,学习文本的边界位置,用于更好的区分文本区域。另外Tian等人[9]提出将同一文本的像素映射到映射空间,在映射空间中令统一文本的映射向量距离相近,不同文本的映射向量距离变远。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/462b5e1472824452a2c530939cda5e59ada226b2d0b745d19dd56068753a7f97\"\n",
"width=\"600\" ></center>\n",
"<br><center>图15 PixelLink框架图</center>\n",
"\n",
"MSR[20]针对文本检测的多尺度问题,提出提取相同图像的多个scale的特征,然后将这些特征融合并上采样到原图尺寸,网络最后预测文本中心区域、文本中心区域每个点到最近的边界点的x坐标偏移和y坐标偏移,最终可以得到文本区域的轮廓坐标集合。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/9597efd68a224d60b74d7c51c99f7ff0ba9939e5cdb84fb79209b7e213f7d039\"\n",
"width=\"600\" ></center>\n",
"<br><center>图16 MSR框架图</center>\n",
" \n",
"针对基于分割的文本算法难以区分相邻文本的问题,PSENet[10]提出渐进式的尺度扩张网络学习文本分割区域,预测不同收缩比例的文本区域,并逐个扩大检测到的文本区域,该方法本质上是边界学习方法的变体,可以有效解决任意形状相邻文本的检测问题。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/fa870b69a2a5423cad7422f64c32e0645dfc31a4ecc94a52832cf8742cded5ba\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图17 PSENet框架图</center>\n",
"\n",
"假设PSENet后处理用了3个不同尺度的kernel,如上图s1,s2,s3所示。首先,从最小kernel s1开始,计算文本分割区域的连通域,得到(b),然后,对连通域沿着上下左右做尺度扩张,对于扩张区域属于s2但不属于s1的像素,进行归类,遇到冲突点时,采用“先到先得”原则,重复尺度扩张的操作,最终可以得到不同文本行的独立的分割区域。\n",
"\n",
"\n",
"Seglink++[17]针对弯曲文本和密集文本问题,提出了一种文本块单元之间的吸引关系和排斥关系的表征,然后设计了一种最小生成树算法进行单元组合得到最终的文本检测框,并提出instance-aware 损失函数使Seglink++方法可以端对端训练。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/1a16568361c0468db537ac25882eed096bca83f9c1544a92aee5239890f9d8d9\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图18 Seglink++框架图</center>\n",
"\n",
"虽然分割方法解决了弯曲文本的检测问题,但是复杂的后处理逻辑以及预测速度也是需要优化的目标。\n",
"\n",
"PAN[11]针对文本检测预测速度慢的问题,从网络设计和后处理方面进行改进,提升算法性能。首先,PAN使用了轻量级的ResNet18作为Backbone,另外设计了轻量级的特征增强模块FPEM和特征融合模块FFM增强Backbone提取的特征。在后处理方面,采用像素聚类方法,沿着预测的文本中心(kernel)四周合并与kernel的距离小于阈值d的像素。PAN保证高精度的同时具有更快的预测速度。\n",
"\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/a76771f91db246ee8be062f96fa2a8abc7598dd87e6d4755b63fac71a4ebc170\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图19 PAN框架图</center>\n",
"\n",
"DBNet[12]针对基于分割的方法需要使用阈值进行二值化处理而导致后处理耗时的问题,提出了可学习阈值并巧妙地设计了一个近似于阶跃函数的二值化函数,使得分割网络在训练的时候能端对端的学习文本分割的阈值。自动调节阈值不仅带来精度的提升,同时简化了后处理,提高了文本检测的性能。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/0d6423e3c79448f8b09090cf2dcf9d0c7baa0f6856c645808502678ae88d2917\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图20 DB框架图</center>\n",
"\n",
"FCENet[16]提出将文本包围曲线用傅立叶变换的参数表示,由于傅里叶系数表示在理论上可以拟合任意的封闭曲线,通过设计合适的模型预测基于傅里叶变换的任意形状文本包围框表示,从而实现了自然场景文本检测中对于高度弯曲文本实例的检测精度的提升。\n",
"\n",
"<center><img src=\"https://ai-studio-static-online.cdn.bcebos.com/45e9a374d97145689a961977f896c8f9f470a66655234c1498e1c8477e277954\"\n",
"width=\"1000\" ></center>\n",
"<br><center>图21 FCENet框架图</center>\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 3. 总结\n",
"\n",
"本节介绍了近几年来文本检测领域的发展,包括基于回归、分割的文本检测方法,并分别列举并介绍了一些经典论文的方法思路。下一节以PaddleOCR开源库为例,详细介绍DBNet的算法原理以及核心代码实现。"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 参考文献\n",
"1. Liao, Minghui, et al. \"Textboxes: A fast text detector with a single deep neural network.\" Thirty-first AAAI conference on artificial intelligence. 2017.\n",
"2. Liao, Minghui, Baoguang Shi, and Xiang Bai. \"Textboxes++: A single-shot oriented scene text detector.\" IEEE transactions on image processing 27.8 (2018): 3676-3690.\n",
"3. Tian, Zhi, et al. \"Detecting text in natural image with connectionist text proposal network.\" European conference on computer vision. Springer, Cham, 2016.\n",
"4. Zhou, Xinyu, et al. \"East: an efficient and accurate scene text detector.\" Proceedings of the IEEE conference on Computer Vision and Pattern Recognition. 2017.\n",
"5. Wang, Fangfang, et al. \"Geometry-aware scene text detection with instance transformation network.\" Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2018.\n",
"6. Yuliang, Liu, et al. \"Detecting curve text in the wild: New dataset and new solution.\" arXiv preprint arXiv:1712.02170 (2017).\n",
"7. Deng, Dan, et al. \"Pixellink: Detecting scene text via instance segmentation.\" Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 32. No. 1. 2018.\n",
"8. Wu, Yue, and Prem Natarajan. \"Self-organized text detection with minimal post-processing via border learning.\" Proceedings of the IEEE International Conference on Computer Vision. 2017.\n",
"9. Tian, Zhuotao, et al. \"Learning shape-aware embedding for scene text detection.\" Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2019.\n",
"10. Wang, Wenhai, et al. \"Shape robust text detection with progressive scale expansion network.\" Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2019.\n",
"11. Wang, Wenhai, et al. \"Efficient and accurate arbitrary-shaped text detection with pixel aggregation network.\" Proceedings of the IEEE/CVF International Conference on Computer Vision. 2019.\n",
"12. Liao, Minghui, et al. \"Real-time scene text detection with differentiable binarization.\" Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 34. No. 07. 2020.\n",
"13. Hochreiter, Sepp, and Jürgen Schmidhuber. \"Long short-term memory.\" Neural computation 9.8 (1997): 1735-1780.\n",
"14. Dai, Pengwen, et al. \"Progressive Contour Regression for Arbitrary-Shape Scene Text Detection.\" Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2021.\n",
"15. He, Minghang, et al. \"MOST: A Multi-Oriented Scene Text Detector with Localization Refinement.\" Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2021.\n",
"16. Zhu, Yiqin, et al. \"Fourier contour embedding for arbitrary-shaped text detection.\" Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2021.\n",
"17. Tang, Jun, et al. \"Seglink++: Detecting dense and arbitrary-shaped scene text by instance-aware component grouping.\" Pattern recognition 96 (2019): 106954.\n",
"18. Wang, Yuxin, et al. \"Contournet: Taking a further step toward accurate arbitrary-shaped scene text detection.\" Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020.\n",
"19. Zhang, Chengquan, et al. \"Look more than once: An accurate detector for text of arbitrary shapes.\" Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2019.\n",
"20. Xue C, Lu S, Zhang W. Msr: Multi-scale shape regression for scene text detection[J]. arXiv preprint arXiv:1901.02596, 2019. \n"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "py35-paddle1.2.0"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.4"
}
},
"nbformat": 4,
"nbformat_minor": 1
}
因为 它太大了无法显示 source diff 。你可以改为 查看blob
{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"\n",
"# 文本识别算法理论\n",
"\n",
"本章主要介绍文本识别算法的理论知识,包括背景介绍、算法分类和部分经典论文思路。\n",
"\n",
"通过本章的学习,你可以掌握:\n",
"\n",
"1. 文本识别的目标\n",
"\n",
"2. 文本识别算法的分类\n",
"\n",
"3. 各类算法的典型思想\n",
"\n",
"\n",
"## 1 背景介绍\n",
"\n",
"文本识别是OCR(Optical Character Recognition)的一个子任务,其任务为识别一个固定区域的的文本内容。在OCR的两阶段方法里,它接在文本检测后面,将图像信息转换为文字信息。\n",
"\n",
"具体地,模型输入一张定位好的文本行,由模型预测出图片中的文字内容和置信度,可视化结果如下图所示:\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/a7c3404f778b489db9c1f686c7d2ff4d63b67c429b454f98b91ade7b89f8e903 width=\"600\"></center>\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/e72b1d6f80c342ac951d092bc8c325149cebb3763ec849ec8a2f54e7c8ad60ca width=\"600\"></center>\n",
"\n",
"\n",
"文本识别的应用场景很多,有文档识别、路标识别、车牌识别、工业编号识别等等,根据实际场景可以把文本识别任务分为两个大类:**规则文本识别**和**不规则文本识别**。\n",
"\n",
"* 规则文本识别:主要指印刷字体、扫描文本等,认为文本大致处在水平线位置\n",
"\n",
"* 不规则文本识别: 往往出现在自然场景中,且由于文本曲率、方向、变形等方面差异巨大,文字往往不在水平位置,存在弯曲、遮挡、模糊等问题。\n",
"\n",
"\n",
"下图展示的是 IC15 和 IC13 的数据样式,它们分别代表了不规则文本和规则文本。可以看出不规则文本往往存在扭曲、模糊、字体差异大等问题,更贴近真实场景,也存在更大的挑战性。\n",
"\n",
"因此目前各大算法都试图在不规则数据集上获得更高的指标。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/bae4fce1370b4751a3779542323d0765a02a44eace7b44d2a87a241c13c6f8cf width=\"400\">\n",
"<br><center>IC15 图片样例(不规则文本)</center>\n",
"<img src=https://ai-studio-static-online.cdn.bcebos.com/b55800d3276f4f5fad170ea1b567eb770177fce226f945fba5d3247a48c15c34 width=\"400\"></center>\n",
"<br><center>IC13 图片样例(规则文本)</center>\n",
"\n",
"\n",
"不同的识别算法在对比能力时,往往也在这两大类公开数据集上比较。对比多个维度上的效果,目前较为通用的英文评估集合分类如下:\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/4d0aada261064031a16816b39a37f2ff6af70dbb57004cb7a106ae6485f14684 width=\"600\"></center>\n",
"\n",
"## 2 文本识别算法分类\n",
"\n",
"在传统的文本识别方法中,任务分为3个步骤,即图像预处理、字符分割和字符识别。需要对特定场景进行建模,一旦场景变化就会失效。面对复杂的文字背景和场景变动,基于深度学习的方法具有更优的表现。\n",
"\n",
"多数现有的识别算法可用如下统一框架表示,算法流程被划分为4个阶段:\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/a2750f4170864f69a3af36fc13db7b606d851f2f467d43cea6fbf3521e65450f)\n",
"\n",
"\n",
"我们整理了主流的算法类别和主要论文,参考下表:\n",
"\n",
"<center>\n",
" \n",
"| 算法类别 | 主要思路 | 主要论文 |\n",
"| -------- | --------------- | -------- |\n",
"| 传统算法 | 滑动窗口、字符提取、动态规划 | - |\n",
"| ctc | 基于ctc的方法,序列不对齐,更快速识别 | CRNN, Rosetta |\n",
"| Attention | 基于attention的方法,应用于非常规文本 | RARE, DAN, PREN |\n",
"| Transformer | 基于transformer的方法 | SRN, NRTR, Master, ABINet |\n",
"| 校正 | 校正模块学习文本边界并校正成水平方向 | RARE, ASTER, SAR | \n",
"| 分割 | 基于分割的方法,提取字符位置再做分类 | Text Scanner, Mask TextSpotter |\n",
" \n",
"</center>\n",
"\n",
"\n",
"### 2.1 规则文本识别\n",
"\n",
"\n",
"文本识别的主流算法有两种,分别是基于 CTC (Conectionist Temporal Classification) 的算法和 Sequence2Sequence 算法,区别主要在解码阶段。\n",
"\n",
"基于 CTC 的算法是将编码产生的序列接入 CTC 进行解码;基于 Sequence2Sequence 的方法则是把序列接入循环神经网络(Recurrent Neural Network, RNN)模块进行循环解码,两种方式都验证有效也是主流的两大做法。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/f64eee66e4a6426f934c1befc3b138629324cf7360c74f72bd6cf3c0de9d49bd width=\"600\"></center>\n",
"<br><center>左:基于CTC的方法,右:基于Sequece2Sequence的方法 </center>\n",
"\n",
"\n",
"#### 2.1.1 基于CTC的算法\n",
"\n",
"基于 CTC 最典型的算法是CRNN (Convolutional Recurrent Neural Network)[1],它的特征提取部分使用主流的卷积结构,常用的有ResNet、MobileNet、VGG等。由于文本识别任务的特殊性,输入数据中存在大量的上下文信息,卷积神经网络的卷积核特性使其更关注于局部信息,缺乏长依赖的建模能力,因此仅使用卷积网络很难挖掘到文本之间的上下文联系。为了解决这一问题,CRNN文本识别算法引入了双向 LSTM(Long Short-Term Memory) 用来增强上下文建模,通过实验证明双向LSTM模块可以有效的提取出图片中的上下文信息。最终将输出的特征序列输入到CTC模块,直接解码序列结果。该结构被验证有效,并广泛应用在文本识别任务中。Rosetta[2]是FaceBook提出的识别网络,由全卷积模型和CTC组成。Gao Y[3]等人使用CNN卷积替代LSTM,参数更少,性能提升精度持平。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/d3c96dd9e9794fddb12fa16f926abdd3485194f0a2b749e792e436037490899b width=\"600\"></center>\n",
"<center> CRNN 结构图 </center>\n",
"\n",
"\n",
"#### 2.1.2 Sequence2Sequence 算法\n",
"\n",
"Sequence2Sequence 算法是由编码器 Encoder 把所有的输入序列都编码成一个统一的语义向量,然后再由解码器Decoder解码。在解码器Decoder解码的过程中,不断地将前一个时刻的输出作为后一个时刻的输入,循环解码,直到输出停止符为止。一般编码器是一个RNN,对于每个输入的词,编码器输出向量和隐藏状态,并将隐藏状态用于下一个输入的单词,循环得到语义向量;解码器是另一个RNN,它接收编码器输出向量并输出一系列字以创建转换。受到 Sequence2Sequence 在翻译领域的启发, Shi[4]提出了一种基于注意的编解码框架来识别文本,通过这种方式,rnn能够从训练数据中学习隐藏在字符串中的字符级语言模型。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/f575333696b7438d919975dc218e61ccda1305b638c5497f92b46a7ec3b85243 width=\"400\" hight=\"500\"></center>\n",
"<center> Sequence2Sequence 结构图 </center>\n",
"\n",
"以上两个算法在规则文本上都有很不错的效果,但由于网络设计的局限性,这类方法很难解决弯曲和旋转的不规则文本识别任务。为了解决这类问题,部分算法研究人员在以上两类算法的基础上提出了一系列改进算法。\n",
"\n",
"### 2.2 不规则文本识别\n",
"\n",
"* 不规则文本识别算法可以被分为4大类:基于校正的方法;基于 Attention 的方法;基于分割的方法;基于 Transformer 的方法。\n",
"\n",
"#### 2.2.1 基于校正的方法\n",
"\n",
"基于校正的方法利用一些视觉变换模块,将非规则的文本尽量转换为规则文本,然后使用常规方法进行识别。\n",
"\n",
"RARE[4]模型首先提出了对不规则文本的校正方案,整个网络分为两个主要部分:一个空间变换网络STN(Spatial Transformer Network) 和一个基于Sequence2Squence的识别网络。其中STN就是校正模块,不规则文本图像进入STN,通过TPS(Thin-Plate-Spline)变换成一个水平方向的图像,该变换可以一定程度上校正弯曲、透射变换的文本,校正后送入序列识别网络进行解码。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/66406f89507245e8a57969b9bed26bfe0227a8cf17a84873902dd4a464b97bb5 width=\"600\"></center>\n",
"<center> RARE 结构图 </center>\n",
"\n",
"RARE论文指出,该方法在不规则文本数据集上有较大的优势,特别比较了CUTE80和SVTP这两个数据集,相较CRNN高出5个百分点以上,证明了校正模块的有效性。基于此[6]同样结合了空间变换网络(STN)和基于注意的序列识别网络的文本识别系统。\n",
"\n",
"基于校正的方法有较好的迁移性,除了RARE这类基于Attention的方法外,STAR-Net[5]将校正模块应用到基于CTC的算法上,相比传统CRNN也有很好的提升。\n",
"\n",
"#### 2.2.2 基于Attention的方法\n",
"\n",
"基于 Attention 的方法主要关注的是序列之间各部分的相关性,该方法最早在机器翻译领域提出,认为在文本翻译的过程中当前词的结果主要由某几个单词影响的,因此需要给有决定性的单词更大的权重。在文本识别领域也是如此,将编码后的序列解码时,每一步都选择恰当的context来生成下一个状态,这样有利于得到更准确的结果。\n",
"\n",
"R^2AM [7] 首次将 Attention 引入文本识别领域,该模型首先将输入图像通过递归卷积层提取编码后的图像特征,然后利用隐式学习到的字符级语言统计信息通过递归神经网络解码输出字符。在解码过程中引入了Attention 机制实现了软特征选择,以更好地利用图像特征,这一有选择性的处理方式更符合人类的直觉。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/a64ef10d4082422c8ac81dcda4ab75bf1db285d6b5fd462a8f309240445654d5 width=\"600\"></center>\n",
"<center> R^2AM 结构图 </center>\n",
"\n",
"后续有大量算法在Attention领域进行探索和更新,例如SAR[8]将1D attention拓展到2D attention上,校正模块提到的RARE也是基于Attention的方法。实验证明基于Attention的方法相比CTC的方法有很好的精度提升。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/4e2507fb58d94ec7a9b4d17151a986c84c5053114e05440cb1e7df423d32cb02 width=\"600\"></center>\n",
"\n",
"\n",
"#### 2.2.3 基于分割的方法\n",
"\n",
"基于分割的方法是将文本行的各字符作为独立个体,相比与对整个文本行做矫正后识别,识别分割出的单个字符更加容易。它试图从输入的文本图像中定位每个字符的位置,并应用字符分类器来获得这些识别结果,将复杂的全局问题简化成了局部问题解决,在不规则文本场景下有比较不错的效果。然而这种方法需要字符级别的标注,数据获取上存在一定的难度。Lyu[9]等人提出了一种用于单词识别的实例分词模型,该模型在其识别部分使用了基于 FCN(Fully Convolutional Network) 的方法。[10]从二维角度考虑文本识别问题,设计了一个字符注意FCN来解决文本识别问题,当文本弯曲或严重扭曲时,该方法对规则文本和非规则文本都具有较优的定位结果。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/fd3e8ef0d6ce4249b01c072de31297ca5d02fc84649846388f890163b624ff10 width=\"800\"></center>\n",
"<center> Mask TextSpotter 结构图 </center>\n",
"\n",
"\n",
"\n",
"#### 2.2.4 基于Transformer的方法\n",
"\n",
"随着 Transformer 的快速发展,分类和检测领域都验证了 Transformer 在视觉任务中的有效性。如规则文本识别部分所说,CNN在长依赖建模上存在局限性,Transformer 结构恰好解决了这一问题,它可以在特征提取器中关注全局信息,并且可以替换额外的上下文建模模块(LSTM)。\n",
"\n",
"一部分文本识别算法使用 Transformer 的 Encoder 结构和卷积共同提取序列特征,Encoder 由多个 MultiHeadAttentionLayer 和 Positionwise Feedforward Layer 堆叠而成的block组成。MulitHeadAttention 中的 self-attention 利用矩阵乘法模拟了RNN的时序计算,打破了RNN中时序长时依赖的障碍。也有一部分算法使用 Transformer 的 Decoder 模块解码,相比传统RNN可获得更强的语义信息,同时并行计算具有更高的效率。\n",
"\n",
"SRN[11] 算法将Transformer的Encoder模块接在ResNet50后,增强了2D视觉特征。并提出了一个并行注意力模块,将读取顺序用作查询,使得计算与时间无关,最终并行输出所有时间步长的对齐视觉特征。此外SRN还利用Transformer的Eecoder作为语义模块,将图片的视觉信息和语义信息做融合,在遮挡、模糊等不规则文本上有较大的收益。\n",
"\n",
"NRTR[12] 使用了完整的Transformer结构对输入图片进行编码和解码,只使用了简单的几个卷积层做高层特征提取,在文本识别上验证了Transformer结构的有效性。\n",
"\n",
"<center><img src=https://ai-studio-static-online.cdn.bcebos.com/e7859f4469a842f0bd450e7e793a679d6e828007544241d09785c9b4ea2424a2 width=\"800\"></center>\n",
"<center> NRTR 结构图 </center>\n",
"\n",
"SRACN[13]使用Transformer的解码器替换LSTM,再一次验证了并行训练的高效性和精度优势。\n",
"\n",
"## 3 总结\n",
"\n",
"本节主要介绍了文本识别相关的理论知识和主流算法,包括基于CTC的方法、基于Sequence2Sequence的方法以及基于分割的方法,并分别列举了经典论文的思路和贡献。下一节将基于CRNN算法进行实践课程讲解,从组网到优化完成整个训练过程,\n",
"\n",
"## 4 参考文献\n",
"\n",
"\n",
"[1]Shi, B., Bai, X., & Yao, C. (2016). An end-to-end trainable neural network for image-based sequence recognition and its application to scene text recognition. IEEE transactions on pattern analysis and machine intelligence, 39(11), 2298-2304.\n",
"\n",
"[2]Fedor Borisyuk, Albert Gordo, and Viswanath Sivakumar. Rosetta: Large scale system for text detection and recognition in images. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pages 71–79. ACM, 2018.\n",
"\n",
"[3]Gao, Y., Chen, Y., Wang, J., & Lu, H. (2017). Reading scene text with attention convolutional sequence modeling. arXiv preprint arXiv:1709.04303.\n",
"\n",
"[4]Shi, B., Wang, X., Lyu, P., Yao, C., & Bai, X. (2016). Robust scene text recognition with automatic rectification. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4168-4176).\n",
"\n",
"[5] Star-Net Max Jaderberg, Karen Simonyan, Andrew Zisserman, et al. Spa- tial transformer networks. In Advances in neural information processing systems, pages 2017–2025, 2015.\n",
"\n",
"[6]Baoguang Shi, Mingkun Yang, XingGang Wang, Pengyuan Lyu, Xiang Bai, and Cong Yao. Aster: An attentional scene text recognizer with flexible rectification. IEEE transactions on pattern analysis and machine intelligence, 31(11):855–868, 2018.\n",
"\n",
"[7] Lee C Y , Osindero S . Recursive Recurrent Nets with Attention Modeling for OCR in the Wild[C]// IEEE Conference on Computer Vision & Pattern Recognition. IEEE, 2016.\n",
"\n",
"[8]Li, H., Wang, P., Shen, C., & Zhang, G. (2019, July). Show, attend and read: A simple and strong baseline for irregular text recognition. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 33, No. 01, pp. 8610-8617).\n",
"\n",
"[9]P. Lyu, C. Yao, W. Wu, S. Yan, and X. Bai. Multi-oriented scene text detection via corner localization and region segmentation. In Proc. CVPR, pages 7553–7563, 2018.\n",
"\n",
"[10] Liao, M., Zhang, J., Wan, Z., Xie, F., Liang, J., Lyu, P., ... & Bai, X. (2019, July). Scene text recognition from two-dimensional perspective. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 33, No. 01, pp. 8714-8721).\n",
"\n",
"[11] Yu, D., Li, X., Zhang, C., Liu, T., Han, J., Liu, J., & Ding, E. (2020). Towards accurate scene text recognition with semantic reasoning networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 12113-12122).\n",
"\n",
"[12] Sheng, F., Chen, Z., & Xu, B. (2019, September). NRTR: A no-recurrence sequence-to-sequence model for scene text recognition. In 2019 International Conference on Document Analysis and Recognition (ICDAR) (pp. 781-786). IEEE.\n",
"\n",
"[13]Yang, L., Wang, P., Li, H., Li, Z., & Zhang, Y. (2020). A holistic representation guided attention network for scene text recognition. Neurocomputing, 414, 67-75.\n",
"\n",
"[14]Wang, T., Zhu, Y., Jin, L., Luo, C., Chen, X., Wu, Y., ... & Cai, M. (2020, April). Decoupled attention network for text recognition. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 34, No. 07, pp. 12216-12224).\n",
"\n",
"[15] Wang, Y., Xie, H., Fang, S., Wang, J., Zhu, S., & Zhang, Y. (2021). From two to one: A new scene text recognizer with visual language modeling network. In Proceedings of the IEEE/CVF International Conference on Computer Vision (pp. 14194-14203).\n",
"\n",
"[16] Fang, S., Xie, H., Wang, Y., Mao, Z., & Zhang, Y. (2021). Read Like Humans: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 7098-7107).\n",
"\n",
"[17] Yan, R., Peng, L., Xiao, S., & Yao, G. (2021). Primitive Representation Learning for Scene Text Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 284-293)."
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "py35-paddle1.2.0"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.4"
}
},
"nbformat": 4,
"nbformat_minor": 1
}
因为 它太大了无法显示 source diff 。你可以改为 查看blob
因为 它太大了无法显示 source diff 。你可以改为 查看blob
因为 它太大了无法显示 source diff 。你可以改为 查看blob
因为 它太大了无法显示 source diff 。你可以改为 查看blob
{
"cells": [
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"# 文档分析技术\n",
"\n",
"本章主要介绍文档分析技术的理论知识,包括背景介绍、算法分类和对应思路。\n",
"\n",
"通过本章的学习,你可以掌握:\n",
"\n",
"1. 版面分析的分类和典型思想\n",
"2. 表格识别的分类和典型思想\n",
"3. 信息提取的分类和典型思想"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"\n",
"作为信息承载工具,文档的不同布局代表了各种不同的信息,如清单和身份证。文档分析是一个从文档中阅读、解释和提取信息的自动化过程。文档分析常包含以下几个研究方向:\n",
"\n",
"1. 版面分析模块: 将每个文档页面划分为不同的内容区域。该模块不仅可用于划定相关区域和不相关区域,还可用于对其识别的内容类型进行分类。\n",
"2. 光学字符识别 (OCR) 模块: 定位并识别文档中存在的所有文本。\n",
"3. 表格识别模块: 将文档里的表格信息进行识别和转换到excel文件中。\n",
"4. 信息提取模块: 借助OCR结果和图像信息来理解和识别文档中表达的特定信息或信息之间的关系。\n",
"\n",
"由于OCR模块在前面的章节中进行了详细的介绍,接下来将针对上面版面分析、表格识别和信息提取三个模块做单独的介绍。对于每一个模块,会介绍该模块的经典或常用方法以及数据集。"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 1. 版面分析\n",
"\n",
"### 1.1 背景介绍\n",
"\n",
"版面分析主要用于文档检索,关键信息提取,内容分类等,其任务主要是对文档图像进行内容分类,内容的类别一般可分为纯文本、标题、表格、图片和列表等。但是文档布局、格式的多样性和复杂性,文档图像质量差,大规模的带标注的数据集的缺少等问题使得版面分析仍然是一个很有挑战性的任务。\n",
"版面分析任务的可视化如下图所示:\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/2510dc76c66c49b8af079f25d08a9dcba726b2ce53d14c8ba5cd9bd57acecf19\" width=\"1000\"/></center>\n",
"<center>图 1:版面分析效果图</center>\n",
"\n",
"现有的解决办法一般是基于目标检测或语义分割的方法,这类方法基将文档中不同的板式当做不同的目标进行检测或分割。\n",
"\n",
"一些代表性论文被划分为上述两个类别中,具体如下表所示:\n",
"\n",
"| 类别 | 主要论文 |\n",
"| ---------------- | -------- |\n",
"| 基于目标检测的方法 | [Visual Detection with Context](https://aclanthology.org/D19-1348.pdf),[Object Detection](https://arxiv.org/pdf/2003.13197v1.pdf),[VSR](https://arxiv.org/pdf/2105.06220v1.pdf)|\n",
"| 基于语义分割的方法 |[Semantic Segmentation](https://arxiv.org/pdf/1911.12170v2.pdf) |\n",
"\n",
"\n",
"### 1.2 基于目标检测的方法 \n",
"\n",
"Soto Carlos[1]在目标检测算法Faster R-CNN的基础上,结合上下文信息并利用文档内容的固有位置信息来提高区域检测性能。Li Kai [2]等人也提出了一种基于目标检测的文档分析方法,通过引入了特征金字塔对齐模块,区域对齐模块,渲染层对齐模块来解决跨域的问题,这三个模块相互补充,并从一般的图像角度和特定的文档图像角度调整域,从而解决了大型标记训练数据集与目标域不同的问题。下图是一个基于目标检测Faster R-CNN算法进行版面分析的流程图。\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/d396e0d6183243898c0961250ee7a49bc536677079fb4ba2ac87c653f5472f01\" width=\"800\"/></center>\n",
"<center>图 2:基于Faster R-CNN的版面分析流程图</center>\n",
"\n",
"### 1.3 基于语义分割的方法 \n",
"\n",
"Sarkar Mausoom[3]等人提出了一种基于先验的分割机制,在非常高的分辨率的图像上训练文档分割模型,解决了过度缩小原始图像导致的密集区域不同结构无法区分进而合并的问题。Zhang Peng[4]等人结合文档中的视觉、语义和关系提出了一个统一的框架VSR(Vision, Semantics and Relations)用于文档布局分析,该框架使用一个双流网络来提取特定模态的视觉和语义特征,并通过自适应聚合模块自适应地融合这些特征,解决了现有基于CV的方法不同模态融合效率低下和布局组件之间缺乏关系建模的局限性。\n",
"\n",
"### 1.4 数据集\n",
"\n",
"虽然现有的方法可以在一定程度上解决版面分析任务,但是该类方法依赖于大量有标记的训练数据。最近也有很多数据集被提出用于文档分析任务。\n",
"\n",
"1. PubLayNet[5]: 该数据集包含50万张文档图像,其中40万用于训练,5万用于验证,5万用于测试,共标记了表格,文本,图像,标题和列表五种形式\n",
"2. HJDataset[6]: 数据集包含2271张文档图像, 除了内容区域的边界框和掩码之外,它还包括布局元素的层次结构和阅读顺序。\n",
"\n",
"PubLayNet数据集样例如下图所示:\n",
"<center class=\"two\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/4b153117c9384f98a0ce5a6c6e7c205a4b1c57e95c894ccb9688cbfc94e68a1c\" width=\"400\"/><img src=\"https://ai-studio-static-online.cdn.bcebos.com/efb9faea39554760b280f9e0e70631d2915399fa97774eecaa44ee84411c4994\" width=\"400\"/>\n",
"</center>\n",
"<center>图 3:PubLayNet样例</center>\n",
"参考文献:\n",
"\n",
"[1]:Soto C, Yoo S. Visual detection with context for document layout analysis[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019: 3464-3470.\n",
"\n",
"[2]:Li K, Wigington C, Tensmeyer C, et al. Cross-domain document object detection: Benchmark suite and method[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2020: 12915-12924.\n",
"\n",
"[3]:Sarkar M, Aggarwal M, Jain A, et al. Document Structure Extraction using Prior based High Resolution Hierarchical Semantic Segmentation[C]//European Conference on Computer Vision. Springer, Cham, 2020: 649-666.\n",
"\n",
"[4]:Zhang P, Li C, Qiao L, et al. VSR: A Unified Framework for Document Layout Analysis combining Vision, Semantics and Relations[J]. arXiv preprint arXiv:2105.06220, 2021.\n",
"\n",
"[5]:Zhong X, Tang J, Yepes A J. Publaynet: largest dataset ever for document layout analysis[C]//2019 International Conference on Document Analysis and Recognition (ICDAR). IEEE, 2019: 1015-1022.\n",
"\n",
"[6]:Li M, Xu Y, Cui L, et al. DocBank: A benchmark dataset for document layout analysis[J]. arXiv preprint arXiv:2006.01038, 2020.\n",
"\n",
"[7]:Shen Z, Zhang K, Dell M. A large dataset of historical japanese documents with complex layouts[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. 2020: 548-549."
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 2. 表格识别\n",
"\n",
"### 2.1 背景介绍\n",
"\n",
"表格是各类文档中常见的页面元素,随着各类文档的爆炸性增长,如何高效地从文档中找到表格并获取内容与结构信息即表格识别,成为了一个亟需解决的问题。表格识别的难点总结如下:\n",
"\n",
"1. 表格种类和样式复杂多样,例如*不同的行列合并,不同的内容文本类型*等。\n",
"2. 文档的样式本身的样式多样。\n",
"3. 拍摄时的光照环境等\n",
"\n",
"表格识别的任务就是将文档里的表格信息转换到excel文件中,任务可视化如下:\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/99faa017e28b4928a408573406870ecaa251b626e0e84ab685e4b6f06f601a5f\" width=\"1600\"/></center>\n",
"\n",
"\n",
"<center>图 4:表格识别示例图,其中左边为原图,右边为表格识别后的结果图,以Excel形式呈现</center>\n",
"\n",
"现有的表格识别算法根据表格结构重建的原理可以分为下面四大类:\n",
"1. 基于启发式规则的方法\n",
"2. 基于CNN的方法\n",
"3. 基于GCN的方法\n",
"4. 基于End to End的方法\n",
"\n",
"一些代表性论文被划分为上述四个类别中,具体如下表所示:\n",
"| 类别 | 思路 | 主要论文 |\n",
"| ---------------- | ---- | -------- |\n",
"|基于启发式规则的方法|人工设计规则,连通域检测分析处理|[T-Rect](https://www.researchgate.net/profile/Andreas-Dengel/publication/249657389_A_Paper-to-HTML_Table_Converting_System/links/0c9605322c9a67274d000000/A-Paper-to-HTML-Table-Converting-System.pdf),[pdf2table](https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.724.7272&rep=rep1&type=pdf)|\n",
"| 基于CNN的方法 | 目标检测,语义分割 | [CascadeTabNet](https://arxiv.org/pdf/2004.12629v2.pdf), [Multi-Type-TD-TSR](https://arxiv.org/pdf/2105.11021v1.pdf), [LGPMA](https://arxiv.org/pdf/2105.06224v2.pdf), [tabstruct-net](https://arxiv.org/pdf/2010.04565v1.pdf), [CDeC-Net](https://arxiv.org/pdf/2008.10831v1.pdf), [TableNet](https://arxiv.org/pdf/2001.01469v1.pdf), [TableSense](https://arxiv.org/pdf/2106.13500v1.pdf), [Deepdesrt](https://www.dfki.de/fileadmin/user_upload/import/9672_PID4966073.pdf), [Deeptabstr](https://www.dfki.de/fileadmin/user_upload/import/10649_DeepTabStR.pdf), [GTE](https://arxiv.org/pdf/2005.00589v2.pdf), [Cycle-CenterNet](https://arxiv.org/pdf/2109.02199v1.pdf), [FCN](https://www.researchgate.net/publication/339027294_Rethinking_Semantic_Segmentation_for_Table_Structure_Recognition_in_Documents)|\n",
"| 基于GCN的方法 | 基于图神经网络,将表格识别看作图重建问题 | [GNN](https://arxiv.org/pdf/1905.13391v2.pdf), [TGRNet](https://arxiv.org/pdf/2106.10598v3.pdf), [GraphTSR](https://arxiv.org/pdf/1908.04729v2.pdf)|\n",
"| 基于End to End的方法 | 利用attention机制 | [Table-Master](https://arxiv.org/pdf/2105.01848v1.pdf)|\n",
"\n",
"### 2.2 基于启发式规则的传统算法\n",
"早期的表格识别研究主要是基于启发式规则的方法。例如由Kieninger[1]等人提出的T-Rect系统使用自底向上的方法对文档图像进行连通域分析,然后按照定义的规则进行合并,得到逻辑文本块。而之后由Yildiz[2]等人提出的pdf2table则是第一个在PDF文档上进行表格识别的方法,它利用了PDF文件的一些特有信息(例如文字、绘制路径等图像文档中难以获取的信息)来协助表格识别。而在最近的工作中,Koci[3]等人将页面中的布局区域表示为图(Graph)的形式,然后使用了Remove and Conquer(RAC)算法从中将表格作为一个子图识别出来。\n",
"\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/66aeedb3f0924d80aee15f185e6799cc687b51fc20b74b98b338ca2ea25be3f3\" width=\"1000\"/></center>\n",
"<center>图 5:启发式算法示意图</center>\n",
"\n",
"### 2.3 基于深度学习CNN的方法\n",
"随着深度学习技术在计算机视觉、自然语言处理、语音处理等领域的飞速发展,研究者将深度学习技术应用到表格识别领域并取得了不错的效果。\n",
"\n",
"Siddiqui Shoaib Ahmed[12]等人在DeepTabStR算法中,将表格结构识别问题表述为对象检测问题,并利用可变形卷积来进更好的进行表格单元格的检测。Raja Sachin[6]等人提出TabStruct-Net将单元格检测和结构识别在视觉上结合起来进行表格结构识别,解决了现有方法由于表格布局发生较大变化而识别错误的问题,但是该方法无法处理行列出现较多空单元格的问题。\n",
"\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/838be28836444bc1835ac30a25613d8b045a1b5aedd44b258499fe9f93dd298f\" width=\"1600\"/></center>\n",
"<center>图 6:基于深度学习CNN的算法示意图</center>\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/4c40dda737bd44b09a533e1b1dd2e4c6a90ceea083bf4238b7f3c7b21087f409\" width=\"1600\"/></center>\n",
"<center>图 7:基于深度学习CNN的算法错误示例</center>\n",
"\n",
"之前的表格结构识别方法一般是从不同粒度(行/列、文本区域)的元素开始处理问题,容易忽略空单元格合并的问题。Qiao Liang[10]等人提出了一个新框架LGPMA,通过掩码重评分策略充分利用来自局部和全局特征的信息,进而可以获得更可靠的对齐单元格区域,最后引入了包括单元格匹配、空单元格搜索和空单元格合并的表格结构复原pipeline来处理表格结构识别问题。\n",
"\n",
"除了以上单独做表格识别的算法外,也有部分方法将表格检测和表格识别在一个模型里完成,Schreiber Sebastian[11]等人提出了DeepDeSRT,通过Faster RCNN进行表格检测,通过FCN语义分割模型用于表格结构行列检测,但是该方法是用两个独立的模型来解决这两个问题。Prasad Devashish[4]等人提出了一种基于端到端深度学习的方法CascadeTabNet,使用Cascade Mask R-CNN HRNet模型同时进行表格检测和结构识别,解决了以往方法使用独立的两个方法处理表格识别问题的不足。Paliwal Shubham[8]等人提出一种新颖的端到端深度多任务架构TableNet,用于表格检测和结构识别,同时在训练期间向TableNet添加额外的空间语义特征,进一步提高了模型性能。Zheng Xinyi[13]等人提出了表格识别的系统框架GTE,利用单元格检测网络来指导表格检测网络的训练,同时提出了一种层次网络和一种新的基于聚类的单元格结构识别算法,该框架可以接入到任何目标检测模型的后面,方便训练不同的表格识别算法。之前的研究主要集中在从扫描的PDF文档中解析具有简单布局的,对齐良好的表格图像,但是现实场景中的表格一般很复杂,可能存在严重变形,弯曲或者遮挡等问题,因此Long Rujiao[14]等人同时构造了一个现实复杂场景下的表格识别数据集WTW,并提出了一种Cycle-CenterNet方法,它利用循环配对模块优化和提出的新配对损失,将离散单元精确地分组到结构化表中,提高了表格识别的性能。\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/a01f714cbe1f42fc9c45c6658317d9d7da2cec9726844f6b9fa75e30cadc9f76\" width=\"1600\"/></center>\n",
"<center>图 8:端到端算法示意图</center>\n",
"\n",
"基于CNN的方法对跨行列的表格无法很好的处理,因此在后续的方法中,分为了两个研究方法来解决表格中跨行列的问题。\n",
"\n",
"### 2.4 基于深度学习GCN的方法\n",
"近些年来,随着图卷积神经网络(Graph Convolutional Network)的兴起,也有一些研究者尝试将图神经网络应用到表格结构识别问题上。Qasim Shah Rukh[20]等人将表格结构识别问题转换为与图神经网络兼容的图问题,并设计了一种新颖的可微架构,该架构既可以利用卷积神经网络提取特征的优点,也可以利用图神经网络顶点之间有效交互的优点,但是该方法只使用了单元格的位置特征,没有利用语义特征。Chi Zewen[19]等人提出了一种新颖的图神经网络GraphTSR,用于PDF文件中的表格结构识别,它以表格中的单元格为输入,然后通过利用图的边和节点相连的特性来预测单元格之间的关系来识别表格结构,一定程度上解决了跨行或者跨列的单元格识别问题。Xue Wenyuan[21]等人将表格结构识别的问题重新表述为表图重建,并提出了一种用于表格结构识别的端到端方法TGRNet,该方法包含单元格检测分支和单元格逻辑位置分支,这两个分支共同预测不同单元格的空间位置和逻辑位置,解决了之前方法没有关注单元格逻辑位置的问题。\n",
"\n",
"GraphTSR表格识别算法示意图:\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/8ff89661142045a8aef54f8a7a2c69b1d243f8269034406a9e66bee2149f730f\" width=\"1600\"/></center>\n",
"<center>图 9:GraphTSR表格识别算法示意图</center>\n",
"\n",
"### 2.5 基于端到端的方法\n",
"\n",
"和其他使用后处理完成表格结构的重建不同,基于端到端的方法直接使用网络完成表格结构的HTML表示输出\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/7865e58a83824facacfaa91bec12ccf834217cb706454dc5a0c165c203db79fb) | ![](https://ai-studio-static-online.cdn.bcebos.com/77d913b1b92f4a349b8f448e08ba78458d687eef4af142678a073830999f3edc))\n",
"---|---\n",
"图 10:端到端方法的输入输出|图 11:Image Caption示例\n",
"\n",
"端到端的方法大多采用Image Caption(看图说话)的Seq2Seq方法来完成表格结构的预测,如一些基于Attention或Transformer的方法。\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/3571280a9c364d3499a062e3edc724294fb5eaef8b38440991941e87f0af0c3b\" width=\"800\"/></center>\n",
"<center>图 12:Seq2Seq示意图</center>\n",
"\n",
"Ye Jiaquan[22]在TableMaster中通过改进基于Transformer的Master文字算法来得到表格结构输出模型。此外,还添加了一个分支进行框的坐标回归,作者并没有在最后一层将模型拆分为两个分支,而是在第一个 Transformer 解码层之后就将序列预测和框回归解耦为两个分支。其网络结构和原始Master网络的对比如下图所示:\n",
"\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/f573709447a848b4ba7c73a2e297f0304caaca57c5c94588aada1f4cd893946c\" width=\"800\"/></center>\n",
"<center>图 13:左:master网络图,右:TableMaster网络图</center>\n",
"\n",
"\n",
"### 2.6 数据集\n",
"\n",
"由于深度学习方法是数据驱动的方法,需要大量的标注数据对模型进行训练,而现有的数据集规模偏小也是一个重要的制约因素,因此也有一些数据集被提出。\n",
"\n",
"1. PubTabNet[16]: 包含568k表格图像和相应的结构化HTML表示。\n",
"2. PubMed Tables(PubTables-1M)[17]:表格结构识别数据集,包含高度详细的结构注释,460,589张pdf图像用于表格检测任务, 947,642张表格图像用于表格识别任务。\n",
"3. TableBank[18]: 表格检测和识别数据集,使用互联网上Word和Latex文档构建了包含417K高质量标注的表格数据。\n",
"4. SciTSR[19]: 表格结构识别数据集,图像大部分从论文中转换而来,其中包含来自PDF文件的15,000个表格及其相应的结构标签。\n",
"5. TabStructDB[12]: 包括1081个表格区域,这些区域用行和列信息密集标记。\n",
"6. WTW[14]: 大规模数据集场景表格检测识别数据集,该数据集包含各种变形,弯曲和遮挡等情况下的表格数据,共包含14,581 张图像。\n",
"\n",
"数据集示例\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/c9763df56e67434f97cd435100d50ded71ba66d9d4f04d7f8f896d613cdf02b0\" /></center>\n",
"<center>图 14:PubTables-1M数据集样例图</center>\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/64de203bbe584642a74f844ac4b61d1ec3c5a38cacb84443ac961fbcc54a66ce\" width=\"600\"/></center>\n",
"<center>图 15:WTW数据集样例图</center>\n",
"\n",
"\n",
"\n",
"参考文献\n",
"\n",
"[1]:Kieninger T, Dengel A. A paper-to-HTML table converting system[C]//Proceedings of document analysis systems (DAS). 1998, 98: 356-365.\n",
"\n",
"[2]:Yildiz B, Kaiser K, Miksch S. pdf2table: A method to extract table information from pdf files[C]//IICAI. 2005: 1773-1785.\n",
"\n",
"[3]:Koci E, Thiele M, Lehner W, et al. Table recognition in spreadsheets via a graph representation[C]//2018 13th IAPR International Workshop on Document Analysis Systems (DAS). IEEE, 2018: 139-144.\n",
"\n",
"[4]:Prasad D, Gadpal A, Kapadni K, et al. CascadeTabNet: An approach for end to end table detection and structure recognition from image-based documents[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. 2020: 572-573.\n",
"\n",
"[5]:Fischer P, Smajic A, Abrami G, et al. Multi-Type-TD-TSR–Extracting Tables from Document Images Using a Multi-stage Pipeline for Table Detection and Table Structure Recognition: From OCR to Structured Table Representations[C]//German Conference on Artificial Intelligence (Künstliche Intelligenz). Springer, Cham, 2021: 95-108.\n",
"\n",
"[6]:Raja S, Mondal A, Jawahar C V. Table structure recognition using top-down and bottom-up cues[C]//European Conference on Computer Vision. Springer, Cham, 2020: 70-86.\n",
"\n",
"[7]:Agarwal M, Mondal A, Jawahar C V. Cdec-net: Composite deformable cascade network for table detection in document images[C]//2020 25th International Conference on Pattern Recognition (ICPR). IEEE, 2021: 9491-9498.\n",
"\n",
"[8]:Paliwal S S, Vishwanath D, Rahul R, et al. Tablenet: Deep learning model for end-to-end table detection and tabular data extraction from scanned document images[C]//2019 International Conference on Document Analysis and Recognition (ICDAR). IEEE, 2019: 128-133.\n",
"\n",
"[9]:Dong H, Liu S, Han S, et al. Tablesense: Spreadsheet table detection with convolutional neural networks[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2019, 33(01): 69-76.\n",
"\n",
"[10]:Qiao L, Li Z, Cheng Z, et al. LGPMA: Complicated Table Structure Recognition with Local and Global Pyramid Mask Alignment[J]. arXiv preprint arXiv:2105.06224, 2021.\n",
"\n",
"[11]:Schreiber S, Agne S, Wolf I, et al. Deepdesrt: Deep learning for detection and structure recognition of tables in document images[C]//2017 14th IAPR international conference on document analysis and recognition (ICDAR). IEEE, 2017, 1: 1162-1167.\n",
"\n",
"[12]:Siddiqui S A, Fateh I A, Rizvi S T R, et al. Deeptabstr: Deep learning based table structure recognition[C]//2019 International Conference on Document Analysis and Recognition (ICDAR). IEEE, 2019: 1403-1409.\n",
"\n",
"[13]:Zheng X, Burdick D, Popa L, et al. Global table extractor (gte): A framework for joint table identification and cell structure recognition using visual context[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 2021: 697-706.\n",
"\n",
"[14]:Long R, Wang W, Xue N, et al. Parsing Table Structures in the Wild[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision. 2021: 944-952.\n",
"\n",
"[15]:Siddiqui S A, Khan P I, Dengel A, et al. Rethinking semantic segmentation for table structure recognition in documents[C]//2019 International Conference on Document Analysis and Recognition (ICDAR). IEEE, 2019: 1397-1402.\n",
"\n",
"[16]:Zhong X, ShafieiBavani E, Jimeno Yepes A. Image-based table recognition: data, model, and evaluation[C]//Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XXI 16. Springer International Publishing, 2020: 564-580.\n",
"\n",
"[17]:Smock B, Pesala R, Abraham R. PubTables-1M: Towards a universal dataset and metrics for training and evaluating table extraction models[J]. arXiv preprint arXiv:2110.00061, 2021.\n",
"\n",
"[18]:Li M, Cui L, Huang S, et al. Tablebank: Table benchmark for image-based table detection and recognition[C]//Proceedings of the 12th Language Resources and Evaluation Conference. 2020: 1918-1925.\n",
"\n",
"[19]:Chi Z, Huang H, Xu H D, et al. Complicated table structure recognition[J]. arXiv preprint arXiv:1908.04729, 2019.\n",
"\n",
"[20]:Qasim S R, Mahmood H, Shafait F. Rethinking table recognition using graph neural networks[C]//2019 International Conference on Document Analysis and Recognition (ICDAR). IEEE, 2019: 142-147.\n",
"\n",
"[21]:Xue W, Yu B, Wang W, et al. TGRNet: A Table Graph Reconstruction Network for Table Structure Recognition[J]. arXiv preprint arXiv:2106.10598, 2021.\n",
"\n",
"[22]:Ye J, Qi X, He Y, et al. PingAn-VCGroup's Solution for ICDAR 2021 Competition on Scientific Literature Parsing Task B: Table Recognition to HTML[J]. arXiv preprint arXiv:2105.01848, 2021.\n"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 3. Document VQA\n",
"\n",
"老板派任务:开发一个身份证识别系统\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/63bbe893465e4f98b3aec80a042758b520d43e1a993a47e39bce1123c2d29b3f\" width=\"1600\"/></center>\n",
"\n",
"\n",
"> 如何选择方案 \n",
"> 1. 文字检测之后用规则来进行信息提取\n",
"> 2. 文字检测之后用规模型来进行信息提取\n",
"> 3. 外包出去\n",
"\n",
"\n",
"### 3.1 背景介绍\n",
"在VQA(Visual Question Answering)任务中,主要针对图像内容进行提问和回答,但是对于文本图像来说,关注的内容是图像中的文字信息,因此这类方法可以分为自然场景的Text-VQA和扫描文档场景的DocVQA,三者的关系如下图所示。\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/a91cfd5152284152b020ca8a396db7a21fd982e3661540d5998cc19c17d84861\" width=\"600\"/></center>\n",
"<center>图 16: VQA层级</center>\n",
"\n",
"VQA,Text-VQA和DocVQA的示例图如下图所示。\n",
"\n",
"|任务类型|VQA | Text-VQA | DocVQA| \n",
"|---|---|---|---|\n",
"|任务描述|针对**图片内容**提出问题|针对**图片上的文字内容**提出问题|针对**文档图像的文字内容**提出问题|\n",
"|示例图片|![vqa](https://ai-studio-static-online.cdn.bcebos.com/fc21b593276247249591231b3373608151ed8ae7787f4d6ba39e8779fdd12201)|![textvqa](https://ai-studio-static-online.cdn.bcebos.com/cd2404edf3bf430b89eb9b2509714499380cd02e4aa74ec39ca6d7aebcf9a559)|![docvqa](https://ai-studio-static-online.cdn.bcebos.com/0eec30a6f91b4f949c56729b856f7ff600d06abee0774642801c070303edfe83)|\n",
"\n",
"DocVQA由于其更加贴近实际应用场景,涌现出了大批学术界和工业界的工作。在常用的场景中,DocVQA里提问的问题都是固定的,比如身份证场景下的问题一般为\n",
"1. 公民身份号码是什么?\n",
"2. 姓名是什么?\n",
"3. 名族是什么?\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/2d2b86468daf47c98be01f44b8d6efa64bc09e43cd764298afb127f19b07aede\" width=\"800\"/></center>\n",
"<center>图 17: 身份证示例</center>\n",
"\n",
"\n",
"基于这样的先验知识,DocVQA的 研究开始偏向Key Information Extraction(KIE)任务,本次我们也主要讨论KIE相关的研究,KIE任务主要从图像中提取所需要的关键信息,如从身份证中提取出姓名和公民身份号码信息。\n",
"\n",
"KIE通常分为两个子任务进行研究\n",
"1. SER: 语义实体识别 (Semantic Entity Recognition),对每一个检测到的文本进行分类,如将其分为姓名,身份证。如下图中的黑色框和红色框。\n",
"2. RE: 关系抽取 (Relation Extraction),对每一个检测到的文本进行分类,如将其分为问题和的答案。然后对每一个问题找到对应的答案。如下图中的红色框和黑色框分别代表问题和答案,黄色线代表问题和答案之间的对应关系。\n",
"\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/899470ba601349fbbc402a4c83e6cdaee08aaa10b5004977b1f684f346ebe31f\" width=\"800\"/></center>\n",
"<center>图 18: SER,RE任务示例</center>\n",
"\n",
"一般的KIE方法基于命名实体识别(Named Entity Recognition,NER)[4]来研究,但是这类方法只利用了图像中的文本信息,缺少对视觉和结构信息的使用,因此精度不高。在此基础上,近几年的方法都开始将视觉和结构信息与文本信息融合到一起,按照对多模态信息进行融合时所采用的的原理可以将这些方法分为下面三种:\n",
"\n",
"1. 基于Grid的方法\n",
"1. 基于Token的方法\n",
"2. 基于GCN的方法\n",
"3. 基于End to End 的方法\n",
"\n",
"一些代表性论文被划分为上述三个类别中,具体如下表所示:\n",
"| 类别 | 思路 | 主要论文 |\n",
"| ---------------- | ---- | -------- |\n",
"| 基于Grid的方法 |在图像上多模态信息的融合(文本,布局,图像)| [Chargrid](https://arxiv.org/pdf/1809.08799) |\n",
"| 基于Token的方法 |利用Bert这类方法进行多模态信息的融合|[LayoutLM](https://arxiv.org/pdf/1912.13318), [LayoutLMv2](https://arxiv.org/pdf/2012.14740), [StrucText](https://arxiv.org/pdf/2108.02923), |\n",
"| 基于GCN的方法 |利用图网络结构进行多模态信息的融合 |[GCN](https://arxiv.org/pdf/1903.11279), [PICK](https://arxiv.org/pdf/2004.07464), [SDMG-R](https://arxiv.org/pdf/2103.14470),[SERA](https://arxiv.org/pdf/2110.09915) |\n",
"| 基于End to End的方法 |将OCR和关键信息提取统一到一个网络 |[Trie](https://arxiv.org/pdf/2005.13118) |\n",
"\n",
"### 3.2 基于Grid的方法\n",
"\n",
"基于Grid的方法在图像层面进行多模态信息的融合。Chargrid[5]首先对图像进行字符级的文字检测和识别,然后通过将类别的one-hot编码填充到对应的字符区域(下图中右图的非黑色部分)内来完成对网络输入的构建,输入最后通过encoder-decoder结构的CNN网络来进行关键信息的坐标检测和类别分类。\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/f248841769ec4312a9015b4befda37bf29db66226431420ca1faad517783875e\" width=\"800\"/></center>\n",
"<center>图 19: Chargrid数据示例</center>\n",
"\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/0682e52e275b4187a0e74f54961a50091fd3a0cdff734e17bedcbc993f6e29f9\" width=\"800\"/></center>\n",
"<center>图 20: Chargrid网络</center>\n",
"\n",
"\n",
"相比于传统的仅基于文本的方法,该方法能够同时利用文本信息和结构信息,因此能够取得一定的精度提升,但是该方法对文本和结构信息的融合只是做了简单的嵌入,并没有很好的将二者进行融合\n",
"\n",
"### 3.3 基于Token的方法\n",
"LayoutLM[6]将2D位置信息和文本信息一起编码到BERT模型中,并且借鉴NLP中Bert的预训练思想,在大规模的数据集上进行预训练,在下游任务中,LayoutLM还加入了图像信息来进一步提升模型性能。LayoutLM虽然将文本,位置和图像信息做了融合,但是图像信息是在下游任务的训练中进行融合,这样对三种信息的多模态融合并不充分。LayoutLMv2[7]在LayoutLM的基础上,通过transformers在预训练阶段将图像信息和文本,layout信息进行融合,还在Transformer中加入空间感知自注意力机制辅助模型更好地融合视觉和文本特征。LayoutLMv2虽然在预训练阶段对文本,位置和图像信息做了融合,但是由于预训练任务的限制,模型学到的视觉特征不够精细。StrucTexT[8]在以往多模态方法的基础上,在预训练任务提出Sentence Length Prediction (SLP) 和Paired Boxes Direction (PBD)两个新任务来帮助网络学习精细的视觉特征,其中SLP任务让模型学习文本段的长度,PDB任务让模型学习Box方向之间的匹配关系。通过这两个新的预训练任务,能够加速文本、视觉和布局信息之间的深度跨模态融合。\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/17a26ade09ee4311b90e49a1c61d88a72a82104478434f9dabd99c27a65d789b) | ![](https://ai-studio-static-online.cdn.bcebos.com/d75addba67ef4b06a02ae40145e609d3692d613ff9b74cec85123335b465b3cc))\n",
"---|---\n",
"图 21:transformer算法流程图|图 22:LayoutLMv2算法流程图\n",
"\n",
"### 3.4 基于GCN的方法\n",
"\n",
"现有的基于GCN的方法[10]虽然利用了文字和结构信息,但是没有对图像信息进行很好的利用。PICK[11]在GCN网络中加入了图像信息并且提出graph learning module来自动学习edge的类型。SDMG-R [12]将图像编码为双模态图,图的节点为文字区域的视觉和文本信息,边表示相邻文本直接的空间关系,通过迭代地沿边传播信息和推理图节点类别,SDMG-R解决了现有的方法对没见过的模板无能为力的问题。\n",
"\n",
"\n",
"PICK流程图如下图所示:\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/d3282959e6b2448c89b762b3b9bbf6197a0364b101214a1f83cf01a28623c01c\" width=\"800\"/></center>\n",
"<center>图 23:PICK算法流程图</center>\n",
"\n",
"SERA[10]将依存句法分析里的biaffine parser引入到文档关系抽取中,并且使用GCN来融合文本和视觉信息。\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/a97b7647968a4fa59e7b14b384dd7ffe812f158db8f741459b6e6bb0e8b657c7\" width=\"800\"/></center>\n",
"<center>图 24:SERA算法流程图</center>\n",
"\n",
"### 3.5 基于End to End 的方法\n",
"\n",
"现有的方法将KIE分为两个独立的任务:文本读取和信息提取,然而他们主要关注于改进信息提取任务,而忽略了文本读取和信息提取是相互关联的,因此,Trie[9]提出了一个统一的端到端网络,可以同时学习这两个任务,并且在学习过程中相互加强。\n",
"\n",
"<center class=\"img\">\n",
"<img src=\"https://ai-studio-static-online.cdn.bcebos.com/6e4a3b0f65254f6b9d40cea0875854d4f47e1dca6b1e408cad435b3629600608\" width=\"1300\"/></center>\n",
"<center>图 25: Trie算法流程图</center>\n",
"\n",
"\n",
"### 3.6 数据集\n",
"用于KIE的数据集主要有下面两个:\n",
"1. SROIE: SROIE数据集[2]的任务3旨在从扫描收据中提取四个预定义的信息:公司、日期、地址或总数。数据集中有626个样本用于训练,347个样本用于测试。\n",
"2. FUNSD: FUNSD数据集[3]是一个用于从扫描文档中提取表单信息的数据集。它包含199个标注好的真实扫描表单。199个样本中149个用于训练,50个用于测试。FUNSD数据集为每个单词分配一个语义实体标签:问题、答案、标题或其他。\n",
"3. XFUN: XFUN数据集是微软提出的一个多语言数据集,包含7种语言,每种语言包含149张训练集,50张测试集。\n",
"\n",
"\n",
"![](https://ai-studio-static-online.cdn.bcebos.com/dfdf530d79504761919c1f093f9a86dac21e6db3304c4892998ea1823f3187c6) | ![](https://ai-studio-static-online.cdn.bcebos.com/3b2a9f9476be4e7f892b73bd7096ce8d88fe98a70bae47e6ab4c5fcc87e83861))\n",
"---|---\n",
"图 26: sroie示例图|图 27: xfun示例图\n",
"\n",
"参考文献:\n",
"\n",
"[1]:Mathew M, Karatzas D, Jawahar C V. Docvqa: A dataset for vqa on document images[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision. 2021: 2200-2209.\n",
"\n",
"[2]:Huang Z, Chen K, He J, et al. Icdar2019 competition on scanned receipt ocr and information extraction[C]//2019 International Conference on Document Analysis and Recognition (ICDAR). IEEE, 2019: 1516-1520.\n",
"\n",
"[3]:Jaume G, Ekenel H K, Thiran J P. Funsd: A dataset for form understanding in noisy scanned documents[C]//2019 International Conference on Document Analysis and Recognition Workshops (ICDARW). IEEE, 2019, 2: 1-6.\n",
"\n",
"[4]:Lample G, Ballesteros M, Subramanian S, et al. Neural architectures for named entity recognition[J]. arXiv preprint arXiv:1603.01360, 2016.\n",
"\n",
"[5]:Katti A R, Reisswig C, Guder C, et al. Chargrid: Towards understanding 2d documents[J]. arXiv preprint arXiv:1809.08799, 2018.\n",
"\n",
"[6]:Xu Y, Li M, Cui L, et al. Layoutlm: Pre-training of text and layout for document image understanding[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2020: 1192-1200.\n",
"\n",
"[7]:Xu Y, Xu Y, Lv T, et al. LayoutLMv2: Multi-modal pre-training for visually-rich document understanding[J]. arXiv preprint arXiv:2012.14740, 2020.\n",
"\n",
"[8]:Li Y, Qian Y, Yu Y, et al. StrucTexT: Structured Text Understanding with Multi-Modal Transformers[C]//Proceedings of the 29th ACM International Conference on Multimedia. 2021: 1912-1920.\n",
"\n",
"[9]:Zhang P, Xu Y, Cheng Z, et al. Trie: End-to-end text reading and information extraction for document understanding[C]//Proceedings of the 28th ACM International Conference on Multimedia. 2020: 1413-1422.\n",
"\n",
"[10]:Liu X, Gao F, Zhang Q, et al. Graph convolution for multimodal information extraction from visually rich documents[J]. arXiv preprint arXiv:1903.11279, 2019.\n",
"\n",
"[11]:Yu W, Lu N, Qi X, et al. Pick: Processing key information extraction from documents using improved graph learning-convolutional networks[C]//2020 25th International Conference on Pattern Recognition (ICPR). IEEE, 2021: 4363-4370.\n",
"\n",
"[12]:Sun H, Kuang Z, Yue X, et al. Spatial Dual-Modality Graph Reasoning for Key Information Extraction[J]. arXiv preprint arXiv:2103.14470, 2021."
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": false
},
"source": [
"## 4. 总结\n",
"本节我们主要介绍了文档分析技术相关的三个子模块的理论知识:版面分析、表格识别和信息提取。下面我们会基于PaddleOCR框架对这表格识别和DOC-VQA进行实战教程的讲解。"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "py35-paddle1.2.0"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.7.4"
}
},
"nbformat": 4,
"nbformat_minor": 1
}
......@@ -234,7 +234,7 @@ python3 train_re.py \
--train_label_path "XFUND/zh_train/xfun_normalize_train.json" \
--eval_data_dir "XFUND/zh_val/image" \
--eval_label_path "XFUND/zh_val/xfun_normalize_val.json" \
--label_map_path 'labels/labels_ser.txt' \
--label_map_path "labels/labels_ser.txt" \
--num_train_epochs 200 \
--eval_steps 10 \
--output_dir "output/re/" \
......@@ -258,7 +258,7 @@ python3 train_re.py \
--train_label_path "XFUND/zh_train/xfun_normalize_train.json" \
--eval_data_dir "XFUND/zh_val/image" \
--eval_label_path "XFUND/zh_val/xfun_normalize_val.json" \
--label_map_path 'labels/labels_ser.txt' \
--label_map_path "labels/labels_ser.txt" \
--num_train_epochs 2 \
--eval_steps 10 \
--output_dir "output/re/" \
......@@ -283,7 +283,7 @@ python3 eval_re.py \
--max_seq_length 512 \
--eval_data_dir "XFUND/zh_val/image" \
--eval_label_path "XFUND/zh_val/xfun_normalize_val.json" \
--label_map_path 'labels/labels_ser.txt' \
--label_map_path "labels/labels_ser.txt" \
--output_dir "output/re/" \
--per_gpu_eval_batch_size 8 \
--num_workers 8 \
......@@ -301,7 +301,7 @@ python3 infer_re.py \
--max_seq_length 512 \
--eval_data_dir "XFUND/zh_val/image" \
--eval_label_path "XFUND/zh_val/xfun_normalize_val.json" \
--label_map_path 'labels/labels_ser.txt' \
--label_map_path "labels/labels_ser.txt" \
--output_dir "output/re/" \
--per_gpu_eval_batch_size 1 \
--seed 2048
......
......@@ -24,7 +24,7 @@ import paddle
from paddlenlp.transformers import LayoutXLMTokenizer, LayoutXLMModel, LayoutXLMForRelationExtraction
from xfun import XFUNDataset
from utils import parse_args, get_bio_label_maps, print_arguments
from vqa_utils import parse_args, get_bio_label_maps, print_arguments
from data_collator import DataCollator
from metric import re_score
......
......@@ -33,7 +33,7 @@ from paddlenlp.transformers import LayoutLMModel, LayoutLMTokenizer, LayoutLMFor
from xfun import XFUNDataset
from losses import SERLoss
from utils import parse_args, get_bio_label_maps, print_arguments
from vqa_utils import parse_args, get_bio_label_maps, print_arguments
from ppocr.utils.logging import get_logger
......
......@@ -15,7 +15,7 @@ import paddle
from paddlenlp.transformers import LayoutXLMTokenizer, LayoutXLMModel, LayoutXLMForRelationExtraction
from xfun import XFUNDataset
from utils import parse_args, get_bio_label_maps, draw_re_results
from vqa_utils import parse_args, get_bio_label_maps, draw_re_results
from data_collator import DataCollator
from ppocr.utils.logging import get_logger
......
......@@ -14,6 +14,10 @@
import os
import sys
__dir__ = os.path.dirname(os.path.abspath(__file__))
sys.path.append(__dir__)
import json
import cv2
import numpy as np
......@@ -22,7 +26,7 @@ from copy import deepcopy
import paddle
# relative reference
from utils import parse_args, get_image_file_list, draw_ser_results, get_bio_label_maps
from vqa_utils import parse_args, get_image_file_list, draw_ser_results, get_bio_label_maps
from paddlenlp.transformers import LayoutXLMModel, LayoutXLMTokenizer, LayoutXLMForTokenClassification
from paddlenlp.transformers import LayoutLMModel, LayoutLMTokenizer, LayoutLMForTokenClassification
......
......@@ -14,6 +14,10 @@
import os
import sys
__dir__ = os.path.dirname(os.path.abspath(__file__))
sys.path.append(__dir__)
import json
import cv2
import numpy as np
......@@ -25,9 +29,9 @@ from paddlenlp.transformers import LayoutXLMModel, LayoutXLMTokenizer, LayoutXLM
from paddlenlp.transformers import LayoutLMModel, LayoutLMTokenizer, LayoutLMForTokenClassification
# relative reference
from utils import parse_args, get_image_file_list, draw_ser_results, get_bio_label_maps
from vqa_utils import parse_args, get_image_file_list, draw_ser_results, get_bio_label_maps
from utils import pad_sentences, split_page, preprocess, postprocess, merge_preds_list_with_ocr_info
from vqa_utils import pad_sentences, split_page, preprocess, postprocess, merge_preds_list_with_ocr_info
MODELS = {
'LayoutXLM':
......
......@@ -24,7 +24,7 @@ import paddle
from paddlenlp.transformers import LayoutXLMModel, LayoutXLMTokenizer, LayoutXLMForRelationExtraction
# relative reference
from utils import parse_args, get_image_file_list, draw_re_results
from vqa_utils import parse_args, get_image_file_list, draw_re_results
from infer_ser_e2e import SerPredictor
......
......@@ -27,7 +27,7 @@ import paddle
from paddlenlp.transformers import LayoutXLMTokenizer, LayoutXLMModel, LayoutXLMForRelationExtraction
from xfun import XFUNDataset
from utils import parse_args, get_bio_label_maps, print_arguments, set_seed
from vqa_utils import parse_args, get_bio_label_maps, print_arguments, set_seed
from data_collator import DataCollator
from eval_re import evaluate
......
......@@ -32,7 +32,7 @@ from paddlenlp.transformers import LayoutXLMModel, LayoutXLMTokenizer, LayoutXLM
from paddlenlp.transformers import LayoutLMModel, LayoutLMTokenizer, LayoutLMForTokenClassification
from xfun import XFUNDataset
from utils import parse_args, get_bio_label_maps, print_arguments, set_seed
from vqa_utils import parse_args, get_bio_label_maps, print_arguments, set_seed
from eval_ser import evaluate
from losses import SERLoss
from ppocr.utils.logging import get_logger
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:null
quant_export:null
fpgm_export:deploy/slim/prune/export_prune_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o
quant_export:null
fpgm_export:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:null
quant_export:null
fpgm_export:deploy/slim/prune/export_prune_model.py -c test_tipc/configs/ch_ppocr_mobile_v2.0_rec_FPGM/rec_chinese_lite_train_v2.0.yml -o
......
......@@ -13,7 +13,7 @@ inference:tools/infer/predict_rec.py
--rec_batch_num:1
--use_tensorrt:False|True
--precision:int8
--det_model_dir:
--rec_model_dir:
--image_dir:./inference/rec_inference
null:null
--benchmark:True
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c test_tipc/configs/ch_ppocr_server_v2.0_det/det_r50_vd_db.yml -o
quant_export:null
fpgm_export:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c configs/det/det_mv3_db.yml -o
quant_export:null
fpgm_export:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c test_tipc/configs/det_mv3_east_v2.0/det_mv3_east.yml -o
quant_export:null
fpgm_export:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c test_tipc/configs/det_mv3_pse_v2.0/det_mv3_pse.yml -o
quant_export:null
fpgm_export:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c configs/det/det_r50_vd_db.yml -o
quant_export:null
fpgm_export:null
......@@ -34,7 +34,7 @@ distill_export:null
export1:null
export2:null
##
train_model:./inference/ch_ppocr_server_v2.0_det_train/best_accuracy
train_model:./inference/det_r50_vd_db_v2.0_train/best_accuracy
infer_export:tools/export_model.py -c configs/det/det_r50_vd_db.yml -o
infer_quant:False
inference:tools/infer/predict_det.py
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c test_tipc/configs/det_r50_vd_east_v2.0/det_r50_vd_east.yml -o
quant_export:null
fpgm_export:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c test_tipc/configs/det_r50_vd_pse_v2.0/det_r50_vd_pse.yml -o
quant_export:null
fpgm_export:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c test_tipc/configs/det_r50_vd_sast_icdar15_v2.0/det_r50_vd_sast_icdar2015.yml -o
quant_export:null
fpgm_export:null
......@@ -43,7 +43,7 @@ inference:tools/infer/predict_det.py
--cpu_threads:1|6
--rec_batch_num:1
--use_tensorrt:False
--precision:fp32|fp16|int8
--precision:fp32|int8
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c test_tipc/configs/det_r50_vd_sast_totaltext_v2.0/det_r50_vd_sast_totaltext.yml -o
quant_export:null
fpgm_export:null
......@@ -43,7 +43,7 @@ inference:tools/infer/predict_det.py
--cpu_threads:1|6
--rec_batch_num:1
--use_tensorrt:False
--precision:fp32|fp16|int8
--precision:fp32|int8
--det_model_dir:
--image_dir:./inference/ch_det_data_50/all-sum-510/
null:null
......
......@@ -26,7 +26,7 @@ null:null
##
===========================infer_params===========================
Global.save_inference_dir:./output/
Global.pretrained_model:
Global.checkpoints:
norm_export:tools/export_model.py -c configs/e2e/e2e_r50_vd_pg.yml -o
quant_export:null
fpgm_export:null
......
# Web 端基础预测功能测试
Web 端主要基于 Jest-Puppeteer 完成 e2e 测试,其中 Puppeteer 操作 Chrome 完成推理流程,Jest 完成测试流程。
>Puppeteer 是一个 Node 库,它提供了一个高级 API 来通过 DevTools 协议控制 Chromium 或 Chrome
>Jest 是一个 JavaScript 测试框架,旨在确保任何 JavaScript 代码的正确性。
#### 环境准备
* 安装 Node(包含 npm ) (https://nodejs.org/zh-cn/download/)
* 确认是否安装成功,在命令行执行
```sh
# 显示所安 node 版本号,即表示成功安装
node -v
```
* 确认 npm 是否安装成成
```sh
# npm 随着 node 一起安装,一般无需额外安装
# 显示所安 npm 版本号,即表示成功安装
npm -v
```
#### 使用
```sh
# web 测试环境准备
bash test_tipc/prepare_js.sh 'js_infer'
# web 推理测试
bash test_tipc/test_inference_js.sh
```
#### 流程设计
###### paddlejs prepare
1. 判断 node, npm 是否安装
2. 下载测试模型,当前检测模型是 ch_PP-OCRv2_det_infer ,识别模型是 ch_PP-OCRv2_rec_infer[1, 3, 32, 320]。如果需要替换模型,可直接将模型文件放在test_tipc/web/models/目录下。
- 文本检测模型:https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar
- 文本识别模型:https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar
- 文本识别模型[1, 3, 32, 320]:https://paddlejs.bj.bcebos.com/models/ch_PP-OCRv2_rec_infer.tar
- 保证较为准确的识别效果,需要将文本识别模型导出为输入shape是[1, 3, 32, 320]的静态模型
3. 转换模型, model.pdmodel model.pdiparams 转换为 model.json chunk.dat(检测模型保存地址:test_tipc/web/models/ch_PP-OCRv2/det,识别模型保存地址:test_tipc/web/models/ch_PP-OCRv2/rec)
4. 安装最新版本 ocr sdk @paddlejs-models/ocr@latest
5. 安装测试环境依赖 puppeteer、jest、jest-puppeteer,如果检查到已经安装,则不会进行二次安装
###### paddlejs infer test
1. Jest 执行 server command:`python3 -m http.server 9811` 开启本地服务
2. 启动 Jest 测试服务,通过 jest-puppeteer 插件完成 chrome 操作,加载 @paddlejs-models/ocr 脚本完成推理流程
3. 测试用例为原图识别后的文本结果与预期文本结果(expect.json)进行对比,测试通过有两个标准:
* 原图识别结果逐字符与预期结果对比,误差不超过 **10个字符**
* 原图识别结果每个文本框字符内容与预期结果进行相似度对比,相似度不小于 0.9(全部一致则相似度为1)。
只有满足上述两个标准,视为测试通过。通过为如下显示:
<img width="600" src="https://user-images.githubusercontent.com/43414102/146406599-80b30c66-f2f8-4f57-a68a-007c479ff0f7.png">
......@@ -179,7 +179,7 @@ elif [ ${MODE} = "whole_infer" ];then
cd ./inference/ && tar xf rec_r34_vd_tps_bilstm_ctc_v2.0_train.tar && cd ../
fi
if [ ${model_name} == "ch_ppocr_server_v2.0_rec" ]; then
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/ch_ppocr_server_v2.0_rec_train.tar --no-check-certificate
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_server_v2.0_rec_train.tar --no-check-certificate
cd ./inference/ && tar xf ch_ppocr_server_v2.0_rec_train.tar && cd ../
fi
if [ ${model_name} == "ch_ppocr_mobile_v2.0_rec" ]; then
......@@ -246,11 +246,15 @@ if [ ${MODE} = "klquant_whole_infer" ]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
cd ./inference && tar xf ch_ppocr_mobile_v2.0_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
fi
if [ ${model_name} = "ch_PPOCRv2_det" ]; then
eval_model_name="ch_PP-OCRv2_det_infer"
if [ ${model_name} = "PPOCRv2_ocr_rec_kl" ]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar --no-check-certificate
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/rec_inference.tar --no-check-certificate
cd ./inference && tar xf rec_inference.tar && tar xf ch_PP-OCRv2_rec_infer.tar && cd ../
fi
if [ ${model_name} = "PPOCRv2_ocr_det_kl" ]; then
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar --no-check-certificate
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
cd ./inference && tar xf ${eval_model_name}.tar && tar xf ch_det_data_50.tar && cd ../
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar --no-check-certificate
cd ./inference && tar xf ch_PP-OCRv2_det_infer.tar && tar xf ch_det_data_50.tar && cd ../
fi
if [ ${model_name} = "ch_ppocr_mobile_v2.0_rec_KL" ]; then
wget -nc -P ./inference/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar --no-check-certificate
......
#!/bin/bash
set -o errexit
set -o nounset
shopt -s extglob
# paddlejs prepare 主要流程
# 1. 判断 node, npm 是否安装
# 2. 下载测试模型,当前检测模型是 ch_PP-OCRv2_det_infer ,识别模型是 ch_PP-OCRv2_rec_infer [1, 3, 32, 320]。如果需要替换模型,可直接将模型文件放在test_tipc/web/models/目录下。
# - 文本检测模型:https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar
# - 文本识别模型:https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tar
# - 文本识别模型[1, 3, 32, 320]:https://paddlejs.bj.bcebos.com/models/ch_PP-OCRv2_rec_infer.tar
# - 保证较为准确的识别效果,需要将文本识别模型导出为输入shape[1, 3, 32, 320]的静态模型
# 3. 转换模型, model.pdmodel model.pdiparams 转换为 model.json chunk.dat(检测模型保存地址:test_tipc/web/models/ch_PP-OCRv2/det,识别模型保存地址:test_tipc/web/models/ch_PP-OCRv2/rec)
# 4. 安装最新版本 ocr sdk @paddlejs-models/ocr@latest
# 5. 安装测试环境依赖 puppeteer、jest、jest-puppeteer,如果检查到已经安装,则不会进行二次安装
# 判断是否安装了node
if ! type node >/dev/null 2>&1; then
echo -e "\033[31m node 未安装 \033[0m"
exit
fi
# 判断是否安装了npm
if ! type npm >/dev/null 2>&1; then
echo -e "\033[31m npm 未安装 \033[0m"
exit
fi
# MODE be 'js_infer'
MODE=$1
# js_infer MODE , load model file and convert model to js_infer
if [ ${MODE} != "js_infer" ];then
echo "Please change mode to 'js_infer'"
exit
fi
# saved_model_name
det_saved_model_name=ch_PP-OCRv2_det_infer
rec_saved_model_name=ch_PP-OCRv2_rec_infer
# model_path
model_path=test_tipc/web/models/
rm -rf $model_path
echo ${model_path}${det_saved_model_name}
echo ${model_path}${rec_saved_model_name}
# download ocr_det inference model
wget -nc -P $model_path https://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tar
cd $model_path && tar xf ch_PP-OCRv2_det_infer.tar && cd ../../../
# download ocr_rec inference model
wget -nc -P $model_path https://paddlejs.bj.bcebos.com/models/ch_PP-OCRv2_rec_infer.tar
cd $model_path && tar xf ch_PP-OCRv2_rec_infer.tar && cd ../../../
MYDIR=`pwd`
echo $MYDIR
pip3 install paddlejsconverter
# convert inference model to web model: model.json、chunk.dat
paddlejsconverter \
--modelPath=$model_path$det_saved_model_name/inference.pdmodel \
--paramPath=$model_path$det_saved_model_name/inference.pdiparams \
--outputDir=$model_path$det_saved_model_name/ \
paddlejsconverter \
--modelPath=$model_path$rec_saved_model_name/inference.pdmodel \
--paramPath=$model_path$rec_saved_model_name/inference.pdiparams \
--outputDir=$model_path$rec_saved_model_name/ \
# always install latest ocr sdk
cd test_tipc/web
echo -e "\033[33m Installing the latest ocr sdk... \033[0m"
npm install @paddlejs-models/ocr@latest
npm info @paddlejs-models/ocr
echo -e "\033[32m The latest ocr sdk installed completely.!~ \033[0m"
# install dependencies
if [ `npm list --dept 0 | grep puppeteer | wc -l` -ne 0 ] && [ `npm list --dept 0 | grep jest | wc -l` -ne 0 ];then
echo -e "\033[32m Dependencies have installed \033[0m"
else
echo -e "\033[33m Installing dependencies ... \033[0m"
npm install jest jest-puppeteer puppeteer
echo -e "\033[32m Dependencies installed completely.!~ \033[0m"
fi
# del package-lock.json
rm package-lock.json
#!/bin/bash
set -o errexit
set -o nounset
cd test_tipc/web
# run ocr test in chrome
./node_modules/.bin/jest --config ./jest.config.js
......@@ -259,7 +259,6 @@ else
env=""
elif [ ${#gpu} -le 1 ];then
env="export CUDA_VISIBLE_DEVICES=${gpu}"
eval ${env}
elif [ ${#gpu} -le 15 ];then
IFS=","
array=(${gpu})
......@@ -280,6 +279,7 @@ else
set_amp_config=" "
fi
for trainer in ${trainer_list[*]}; do
eval ${env}
flag_quant=False
if [ ${trainer} = ${pact_key} ]; then
run_train=${pact_trainer}
......@@ -332,7 +332,6 @@ else
cmd="${python} -m paddle.distributed.launch --ips=${ips} --gpus=${gpu} ${run_train} ${set_use_gpu} ${set_save_model} ${set_pretrain} ${set_epoch} ${set_autocast} ${set_batchsize} ${set_train_params1} ${set_amp_config}"
fi
# run train
eval "unset CUDA_VISIBLE_DEVICES"
eval $cmd
status_check $? "${cmd}" "${status_log}"
......
{
"text": [
"纯臻营养护发素",
"产品信息/参数",
"(45元/每公斤,100公斤起订)",
"每瓶22元,1000瓶起订)",
"【品牌】:代加工方式/OEMODM",
"【品名】:纯臻营养护发素",
"【产品编号】:YM-X-3011",
"ODMOEM",
"【净含量】:220ml",
"【适用人群】:适合所有肤质",
"【主要成分】:鲸蜡硬脂醇、燕麦β-葡聚",
"糖、椰油酰胺丙基甜菜碱、泛醌",
"(成品包材)",
"【主要功能】:可紧致头发磷层,从而达到",
"即时持久改善头发光泽的效果,给干燥的头",
"发足够的滋养"
]
}
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta http-equiv="X-UA-Compatible" content="ie=edge">
<title>ocr test</title>
</head>
<body>
<img id="ocr" src="./test.jpg" />
</body>
<script src="./node_modules/@paddlejs-models/ocr/lib/index.js"></script>
</html>
\ No newline at end of file
const expectData = require('./expect.json');
describe('e2e test ocr model', () => {
beforeAll(async () => {
await page.goto(PATH);
});
it('ocr infer and diff test', async () => {
page.on('console', msg => console.log('PAGE LOG:', msg.text()));
const text = await page.evaluate(async () => {
const $ocr = document.querySelector('#ocr');
const ocr = paddlejs['ocr'];
await ocr.init('./models/ch_PP-OCRv2_det_infer', './models/ch_PP-OCRv2_rec_infer');
const res = await ocr.recognize($ocr);
return res.text;
});
// 模型文字识别结果与预期结果diff的字符数
let diffNum = 0;
// 文本框字符串相似度
let similarity = 0;
// 预期字符diff数
const expectedDiffNum = 10;
// 预期文本框字符串相似度
const expecteSimilarity = 0.9;
// 预期文本内容
const expectResult = expectData.text;
expectResult && expectResult.forEach((item, index) => {
const word = text[index];
// 逐字符对比
for(let i = 0; i < item.length; i++) {
if (item[i] !== word[i]) {
console.log('expect: ', item[i], ' word: ', word[i]);
diffNum++;
}
}
// 文本框字符串相似度对比
const s = similar(item, word);
similarity += s;
});
similarity = similarity / expectResult.length;
expect(diffNum).toBeLessThanOrEqual(expectedDiffNum);
expect(similarity).toBeGreaterThanOrEqual(expecteSimilarity);
function similar(string, expect) {
if (!string || !expect) {
return 0;
}
const length = string.length > expect.length ? string.length : expect.length;
const n = string.length;
const m = expect.length;
let data = [];
const min = (a, b, c) => {
return a < b ? (a < c ? a : c) : (b < c ? b : c);
};
let i, j, si, ej, cost;
if (n === 0) return m;
if (m === 0) return n;
for (i = 0; i <= n; i++) {
data[i] = [];
[i][0] = i
}
for (j = 0; j <= m; j++) {
data[0][j] = j;
}
for (i = 1; i <= n; i++) {
si = string.charAt(i - 1);
for (j = 1; j <= m; j++) {
ej = expect.charAt(j - 1);
cost = si === ej ? 0 : 1;
data[i][j] = min(data[i - 1][j] + 1, data[i][j - 1] + 1, data[i - 1][j - 1] + cost);
}
}
return (1 - data[n][m] / length);
}
});
});
// jest-puppeteer.config.js
module.exports = {
launch: {
headless: false,
product: 'chrome'
},
browserContext: 'default',
server: {
command: 'python3 -m http.server 9811',
port: 9811,
launchTimeout: 10000,
debug: true
}
};
// For a detailed explanation regarding each configuration property and type check, visit:
// https://jestjs.io/docs/en/configuration.html
module.exports = {
preset: 'jest-puppeteer',
// All imported modules in your tests should be mocked automatically
// automock: false,
// Automatically clear mock calls and instances between every test
clearMocks: true,
// An object that configures minimum threshold enforcement for coverage results
// coverageThreshold: undefined,
// A set of global variables that need to be available in all test environments
globals: {
PATH: 'http://localhost:9811'
},
// The maximum amount of workers used to run your tests. Can be specified as % or a number. E.g. maxWorkers: 10% will use 10% of your CPU amount + 1 as the maximum worker number. maxWorkers: 2 will use a maximum of 2 workers.
// maxWorkers: "50%",
// An array of directory names to be searched recursively up from the requiring module's location
// moduleDirectories: [
// "node_modules"
// ],
// An array of file extensions your modules use
moduleFileExtensions: [
'js',
'json',
'jsx',
'ts',
'tsx',
'node'
],
// The root directory that Jest should scan for tests and modules within
// rootDir: undefined,
// A list of paths to directories that Jest should use to search for files in
roots: [
'<rootDir>'
],
// Allows you to use a custom runner instead of Jest's default test runner
// runner: "jest-runner",
// The paths to modules that run some code to configure or set up the testing environment before each test
// setupFiles: [],
// A list of paths to modules that run some code to configure or set up the testing framework before each test
// setupFilesAfterEnv: [],
// The number of seconds after which a test is considered as slow and reported as such in the results.
// slowTestThreshold: 5,
// A list of paths to snapshot serializer modules Jest should use for snapshot testing
// snapshotSerializers: [],
// The test environment that will be used for testing
// testEnvironment: 'jsdom',
// Options that will be passed to the testEnvironment
// testEnvironmentOptions: {},
// An array of regexp pattern strings that are matched against all test paths, matched tests are skipped
testPathIgnorePatterns: [
'/node_modules/'
],
// The regexp pattern or array of patterns that Jest uses to detect test files
testRegex: '.(.+)\\.test\\.(js|ts)$',
// This option allows the use of a custom results processor
// testResultsProcessor: undefined,
// This option allows use of a custom test runner
// testRunner: "jest-circus/runner",
// This option sets the URL for the jsdom environment. It is reflected in properties such as location.href
testURL: 'http://localhost:9898/',
// Setting this value to "fake" allows the use of fake timers for functions such as "setTimeout"
// timers: "real",
// A map from regular expressions to paths to transformers
transform: {
'^.+\\.js$': 'babel-jest'
},
// An array of regexp pattern strings that are matched against all source file paths, matched files will skip transformation
transformIgnorePatterns: [
'/node_modules/',
'\\.pnp\\.[^\\/]+$'
],
// An array of regexp pattern strings that are matched against all modules before the module loader will automatically return a mock for them
// unmockedModulePathPatterns: undefined,
// Indicates whether each individual test should be reported during the run
verbose: true,
// An array of regexp patterns that are matched against all source file paths before re-running tests in watch mode
// watchPathIgnorePatterns: [],
// Whether to use watchman for file crawling
// watchman: true,
testTimeout: 50000
};
......@@ -15,6 +15,7 @@
import argparse
import os
import sys
import platform
import cv2
import numpy as np
import paddle
......@@ -313,6 +314,10 @@ def create_predictor(args, mode, logger):
def get_infer_gpuid():
sysstr = platform.system()
if sysstr == "Windows":
return 0
if not paddle.fluid.core.is_compiled_with_rocm():
cmd = "env | grep CUDA_VISIBLE_DEVICES"
else:
......
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