Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
wangsen
paddle_dbnet
Commits
0888a4ae
Unverified
Commit
0888a4ae
authored
Dec 15, 2020
by
dyning
Committed by
GitHub
Dec 15, 2020
Browse files
Merge pull request #1440 from weisy11/dygraph
Update Styletext doc
parents
a26968a4
e27c136a
Changes
40
Show whitespace changes
Inline
Side-by-side
Showing
20 changed files
with
22 additions
and
10 deletions
+22
-10
StyleTextRec/README.md
StyleTextRec/README.md
+1
-1
StyleTextRec/README_ch.md
StyleTextRec/README_ch.md
+21
-9
StyleTextRec/__init__.py
StyleTextRec/__init__.py
+0
-0
StyleTextRec/arch/__init__.py
StyleTextRec/arch/__init__.py
+0
-0
StyleTextRec/arch/base_module.py
StyleTextRec/arch/base_module.py
+0
-0
StyleTextRec/arch/decoder.py
StyleTextRec/arch/decoder.py
+0
-0
StyleTextRec/arch/encoder.py
StyleTextRec/arch/encoder.py
+0
-0
StyleTextRec/arch/spectral_norm.py
StyleTextRec/arch/spectral_norm.py
+0
-0
StyleTextRec/arch/style_text_rec.py
StyleTextRec/arch/style_text_rec.py
+0
-0
StyleTextRec/configs/config.yml
StyleTextRec/configs/config.yml
+0
-0
StyleTextRec/configs/dataset_config.yml
StyleTextRec/configs/dataset_config.yml
+0
-0
StyleTextRec/doc/images/1.png
StyleTextRec/doc/images/1.png
+0
-0
StyleTextRec/doc/images/2.png
StyleTextRec/doc/images/2.png
+0
-0
StyleTextRec/doc/images/3.png
StyleTextRec/doc/images/3.png
+0
-0
StyleTextRec/doc/images/4.jpg
StyleTextRec/doc/images/4.jpg
+0
-0
StyleTextRec/doc/images/5.png
StyleTextRec/doc/images/5.png
+0
-0
StyleTextRec/doc/images/6.png
StyleTextRec/doc/images/6.png
+0
-0
StyleTextRec/engine/__init__.py
StyleTextRec/engine/__init__.py
+0
-0
StyleTextRec/engine/corpus_generators.py
StyleTextRec/engine/corpus_generators.py
+0
-0
StyleTextRec/engine/predictors.py
StyleTextRec/engine/predictors.py
+0
-0
No files found.
doc/doc_en/s
tyle
_t
ext
_rec_en
.md
→
S
tyle
T
ext
Rec/README
.md
View file @
0888a4ae
...
...
@@ -7,7 +7,7 @@ The following figure shows some example results. In addition, the actual `namepl
#### Preparation
1.
Please refer the
[
QUICK INSTALLATION
](
./installation_en.md
)
to install PaddlePaddle. Python3 environment is strongly recommended.
1.
Please refer the
[
QUICK INSTALLATION
](
.
./doc/doc_en
/installation_en.md
)
to install PaddlePaddle. Python3 environment is strongly recommended.
2.
Download the pretrained models and unzip:
```
bash
...
...
doc/doc_ch/s
tyle
_t
ext
_rec
.md
→
S
tyle
T
ext
Rec/README_ch
.md
View file @
0888a4ae
...
...
@@ -9,21 +9,21 @@
### 工具简介
<div
align=
"center"
>
<img
src=
"
../imgs_style_text
/3.png"
width=
"800"
>
<img
src=
"
doc/images
/3.png"
width=
"800"
>
</div>
Style-Text是对百度自研文本编辑算法《Editing Text in the Wild》中提出的SRNet网络的改进,不同于常用的GAN的方法只选择一个分支,该工具将文本合成任务分解为三个子模块,文本风格迁移模块、背景抽取模块和前背景融合模块,来提升合成数据的效果。下图显示了一些示例结果。
<div
align=
"center"
>
<img
src=
"
../imgs_style_text
/1.png"
width=
"800"
>
<img
src=
"
../imgs_style_text
/2.png"
width=
"800"
>
<img
src=
"
doc/images
/1.png"
width=
"800"
>
<img
src=
"
doc/images
/2.png"
width=
"800"
>
</div>
此外,在实际铭牌文本识别场景和韩语文本识别场景,验证了该合成工具的有效性。
### 环境配置
1.
参考
[
快速安装
](
./installation.md
)
,安装PaddleOCR。强烈建议您使用python3环境。
1.
参考
[
快速安装
](
.
./doc/doc_ch
/installation.md
)
,安装PaddleOCR。强烈建议您使用python3环境。
2.
进入
`style_text_rec`
目录,下载模型,并解压:
```
bash
...
...
@@ -55,7 +55,7 @@ python3 -m tools.synth_image -c configs/config.yml
1.
运行后,会生成
`fake_busion.jpg`
,即为最终结果。
<div
align=
"center"
>
<img
src=
"
../imgs_style_text
/4.jpg"
width=
"300"
>
<img
src=
"
doc/images
/4.jpg"
width=
"300"
>
</div>
除此之外,程序还会生成并保存中间结果:
*
`fake_bg.jpg`
:为风格参考图去掉文字后的背景;
...
...
@@ -81,12 +81,15 @@ python3 -m tools.synth_image -c configs/config.yml --style_image examples/style_
*
`image_home`
:风格图片目录;
*
`label_file`
:风格图片路径列表文件,如果所用数据集有label,则label_file为label文件路径;
*
`with_label`
:标志
`label_file`
是否为label文件。
我们提供了一批
[
样例图
](
https://paddleocr.bj.bcebos.com/dygraph_v2.0/style_text/chkoen_5w.tar
)
供您试用。
*
`CorpusGenerator`
:
*
`method`
:语料生成方法,目前有
`FileCorpus`
和
`EnNumCorpus`
可选。如果使用
`EnNumCorpus`
,则不需要填写其他配置,否则需要修改
`corpus_file`
和
`language`
;
*
`language`
:语料的语种;
*
`corpus_file`
: 语料文件路径。
我们提供了一批
[
样例图
](
https://paddleocr.bj.bcebos.com/dygraph_v2.0/style_text/chkoen_5w.tar
)
供您试用,下面给出了一些示例:
<div
align=
"center"
>
<img
src=
"doc/images/5.png"
width=
"800"
>
</div>
2.
运行
`tools/synth_dataset`
合成数据:
```
bash
...
...
@@ -104,8 +107,17 @@ python3 -m tools.synth_image -c configs/config.yml --style_image examples/style_
### 应用示例
在完成上述操作后,即可得到用于OCR识别的合成数据集,下面给出了一些数据集生成的示例:
<div
align=
"center"
>
<img
src=
"doc/images/6.png"
width=
"800"
>
</div>
请您参考
[
OCR识别文档
](
https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/recognition.md#%E5%90%AF%E5%8A%A8%E8%AE%AD%E7%BB%83
)
,完成训练。
下面展示了一些使用合成数据训练的效果:
接下来请参考
[
OCR识别文档
](
https://github.com/PaddlePaddle/PaddleOCR/blob/dygraph/doc/doc_ch/recognition.md#%E5%90%AF%E5%8A%A8%E8%AE%AD%E7%BB%83
)
,完成训练。
| 场景 | 字符 | 原始数据 | 测试数据 | 只使用原始数据的识别准确率 | 新增合成数据 | 使用合成数据识别准确率 | 指标提升 |
| -------- | ---------- | -------- | -------- | -------------------------- | ------------ | ---------------------- | -------- |
| 金属表面 | 英文和数字 | 2203 | 650 | 0.5938 | 20000 | 0.7546 | 16% |
| 随机背景 | 韩语 | 5631 | 1230 | 0.3012 | 100000 | 0.5057 | 20% |
### 项目结构
```
...
...
s
tyle
_t
ext
_r
ec/__init__.py
→
S
tyle
T
ext
R
ec/__init__.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/arch/__init__.py
→
S
tyle
T
ext
R
ec/arch/__init__.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/arch/base_module.py
→
S
tyle
T
ext
R
ec/arch/base_module.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/arch/decoder.py
→
S
tyle
T
ext
R
ec/arch/decoder.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/arch/encoder.py
→
S
tyle
T
ext
R
ec/arch/encoder.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/arch/spectral_norm.py
→
S
tyle
T
ext
R
ec/arch/spectral_norm.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/arch/style_text_rec.py
→
S
tyle
T
ext
R
ec/arch/style_text_rec.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/configs/config.yml
→
S
tyle
T
ext
R
ec/configs/config.yml
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/configs/dataset_config.yml
→
S
tyle
T
ext
R
ec/configs/dataset_config.yml
View file @
0888a4ae
File moved
doc/imgs_style_text
/1.png
→
StyleTextRec/doc/images
/1.png
View file @
0888a4ae
File moved
doc/imgs_style_text
/2.png
→
StyleTextRec/doc/images
/2.png
View file @
0888a4ae
File moved
doc/imgs_style_text
/3.png
→
StyleTextRec/doc/images
/3.png
View file @
0888a4ae
File moved
doc/imgs_style_text
/4.jpg
→
StyleTextRec/doc/images
/4.jpg
View file @
0888a4ae
File moved
doc/imgs_style_text
/5.png
→
StyleTextRec/doc/images
/5.png
View file @
0888a4ae
File moved
StyleTextRec/doc/images/6.png
0 → 100644
View file @
0888a4ae
125 KB
s
tyle
_t
ext
_r
ec/engine/__init__.py
→
S
tyle
T
ext
R
ec/engine/__init__.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/engine/corpus_generators.py
→
S
tyle
T
ext
R
ec/engine/corpus_generators.py
View file @
0888a4ae
File moved
s
tyle
_t
ext
_r
ec/engine/predictors.py
→
S
tyle
T
ext
R
ec/engine/predictors.py
View file @
0888a4ae
File moved
Prev
1
2
Next
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment