Skip to content

GitLab

Commit aa59fca5 authored by Leif's avatar Leif
Browse files

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

parents 12d15752 f01f24c7
Hide whitespace changes
Inline Side-by-side
Showing with 412 additions and 24 deletions
deploy/slim/quantization/README.md
View file @ aa59fca5
......@@ -42,7 +42,7 @@ python deploy/slim/quantization/quant.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3
# 比如下载提供的训练模型
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar
tar -xf ch_ppocr_mobile_v2.0_det_train.tar
python deploy/slim/quantization/quant.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_model_dir=./output/quant_inference_model
python deploy/slim/quantization/quant.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_model_dir=./output/quant_model
```
如果要训练识别模型的量化，修改配置文件和加载的模型参数即可。
......
deploy/slim/quantization/export_model.py
View file @ aa59fca5
......@@ -35,17 +35,7 @@ from ppocr.metrics import build_metric
import tools.program as program
from paddleslim.dygraph.quant import QAT
from ppocr.data import build_dataloader
def export_single_model(quanter, model, infer_shape, save_path, logger):
quanter.save_quantized_model(
model,
save_path,
input_spec=[
paddle.static.InputSpec(
shape=[None] + infer_shape, dtype='float32')
])
logger.info('inference QAT model is saved to {}'.format(save_path))
from tools.export_model import export_single_model
def main():
......@@ -84,17 +74,54 @@ def main():
config['Global'])
# build model
# for rec algorithm
if hasattr(post_process_class, 'character'):
char_num = len(getattr(post_process_class, 'character'))
if config['Architecture']["algorithm"] in ["Distillation",
]: # distillation model
for key in config['Architecture']["Models"]:
config['Architecture']["Models"][key]["Head"][
'out_channels'] = char_num
if config['Architecture']['Models'][key]['Head'][
'name'] == 'MultiHead': # for multi head
if config['PostProcess'][
'name'] == 'DistillationSARLabelDecode':
char_num = char_num - 2
# update SARLoss params
assert list(config['Loss']['loss_config_list'][-1].keys())[
0] == 'DistillationSARLoss'
config['Loss']['loss_config_list'][-1][
'DistillationSARLoss']['ignore_index'] = char_num + 1
out_channels_list = {}
out_channels_list['CTCLabelDecode'] = char_num
out_channels_list['SARLabelDecode'] = char_num + 2
config['Architecture']['Models'][key]['Head'][
'out_channels_list'] = out_channels_list
else:
config['Architecture']["Models"][key]["Head"][
'out_channels'] = char_num
elif config['Architecture']['Head'][
'name'] == 'MultiHead': # for multi head
if config['PostProcess']['name'] == 'SARLabelDecode':
char_num = char_num - 2
# update SARLoss params
assert list(config['Loss']['loss_config_list'][1].keys())[
0] == 'SARLoss'
if config['Loss']['loss_config_list'][1]['SARLoss'] is None:
config['Loss']['loss_config_list'][1]['SARLoss'] = {
'ignore_index': char_num + 1
}
else:
config['Loss']['loss_config_list'][1]['SARLoss'][
'ignore_index'] = char_num + 1
out_channels_list = {}
out_channels_list['CTCLabelDecode'] = char_num
out_channels_list['SARLabelDecode'] = char_num + 2
config['Architecture']['Head'][
'out_channels_list'] = out_channels_list
else: # base rec model
config['Architecture']["Head"]['out_channels'] = char_num
if config['PostProcess']['name'] == 'SARLabelDecode': # for SAR model
config['Loss']['ignore_index'] = char_num - 1
model = build_model(config['Architecture'])
# get QAT model
......@@ -120,21 +147,22 @@ def main():
for k, v in metric.items():
logger.info('{}:{}'.format(k, v))
infer_shape = [3, 32, 100] if model_type == "rec" else [3, 640, 640]
save_path = config["Global"]["save_inference_dir"]
arch_config = config["Architecture"]
arch_config = config["Architecture"]
if arch_config["algorithm"] in ["Distillation", ]: # distillation model
archs = list(arch_config["Models"].values())
for idx, name in enumerate(model.model_name_list):
model.model_list[idx].eval()
sub_model_save_path = os.path.join(save_path, name, "inference")
export_single_model(quanter, model.model_list[idx], infer_shape,
sub_model_save_path, logger)
export_single_model(model.model_list[idx], archs[idx],
sub_model_save_path, logger, quanter)
else:
save_path = os.path.join(save_path, "inference")
model.eval()
export_single_model(quanter, model, infer_shape, save_path, logger)
export_single_model(model, arch_config, save_path, logger, quanter)
if __name__ == "__main__":
......
deploy/slim/quantization/quant.py
View file @ aa59fca5
......@@ -112,10 +112,48 @@ def main(config, device, logger, vdl_writer):
if config['Architecture']["algorithm"] in ["Distillation",
]: # distillation model
for key in config['Architecture']["Models"]:
config['Architecture']["Models"][key]["Head"][
'out_channels'] = char_num
if config['Architecture']['Models'][key]['Head'][
'name'] == 'MultiHead': # for multi head
if config['PostProcess'][
'name'] == 'DistillationSARLabelDecode':
char_num = char_num - 2
# update SARLoss params
assert list(config['Loss']['loss_config_list'][-1].keys())[
0] == 'DistillationSARLoss'
config['Loss']['loss_config_list'][-1][
'DistillationSARLoss']['ignore_index'] = char_num + 1
out_channels_list = {}
out_channels_list['CTCLabelDecode'] = char_num
out_channels_list['SARLabelDecode'] = char_num + 2
config['Architecture']['Models'][key]['Head'][
'out_channels_list'] = out_channels_list
else:
config['Architecture']["Models"][key]["Head"][
'out_channels'] = char_num
elif config['Architecture']['Head'][
'name'] == 'MultiHead': # for multi head
if config['PostProcess']['name'] == 'SARLabelDecode':
char_num = char_num - 2
# update SARLoss params
assert list(config['Loss']['loss_config_list'][1].keys())[
0] == 'SARLoss'
if config['Loss']['loss_config_list'][1]['SARLoss'] is None:
config['Loss']['loss_config_list'][1]['SARLoss'] = {
'ignore_index': char_num + 1
}
else:
config['Loss']['loss_config_list'][1]['SARLoss'][
'ignore_index'] = char_num + 1
out_channels_list = {}
out_channels_list['CTCLabelDecode'] = char_num
out_channels_list['SARLabelDecode'] = char_num + 2
config['Architecture']['Head'][
'out_channels_list'] = out_channels_list
else: # base rec model
config['Architecture']["Head"]['out_channels'] = char_num
if config['PostProcess']['name'] == 'SARLabelDecode': # for SAR model
config['Loss']['ignore_index'] = char_num - 1
model = build_model(config['Architecture'])
pre_best_model_dict = dict()
......@@ -137,7 +175,7 @@ def main(config, device, logger, vdl_writer):
config['Optimizer'],
epochs=config['Global']['epoch_num'],
step_each_epoch=len(train_dataloader),
parameters=model.parameters())
model=model)
# resume PACT training process
if config["Global"]["checkpoints"] is not None:
......
doc/doc_ch/algorithm.md 0 → 100644
View file @ aa59fca5
# 前沿算法与模型
PaddleOCR将**持续新增**支持OCR领域前沿算法与模型，已支持的模型与使用教程可点击下方列表查看：
- [文本检测算法](./algorithm_overview.md#11-%E6%96%87%E6%9C%AC%E6%A3%80%E6%B5%8B%E7%AE%97%E6%B3%95)
- [文本识别算法](./algorithm_overview.md#12-%E6%96%87%E6%9C%AC%E8%AF%86%E5%88%AB%E7%AE%97%E6%B3%95)
- [端到端算法](./algorithm_overview.md#2-%E6%96%87%E6%9C%AC%E8%AF%86%E5%88%AB%E7%AE%97%E6%B3%95)
**欢迎广大开发者合作共建，贡献更多算法，合入有奖🎁！具体可查看[社区常规赛](https://github.com/PaddlePaddle/PaddleOCR/issues/4982)。**
新增算法可参考如下教程：
- [使用PaddleOCR架构添加新算法](./add_new_algorithm.md)
\ No newline at end of file
doc/doc_ch/algorithm_det_db.md 0 → 100644
View file @ aa59fca5
# DB
- [1. 算法简介](#1)
- [2. 环境配置](#2)
- [3. 模型训练、评估、预测](#3)
- [3.1 训练](#3-1)
- [3.2 评估](#3-2)
- [3.3 预测](#3-3)
- [4. 推理部署](#4)
- [4.1 Python推理](#4-1)
- [4.2 C++推理](#4-2)
- [4.3 Serving服务化部署](#4-3)
- [4.4 更多推理部署](#4-4)
- [5. FAQ](#5)
<a name="1"></a>
## 1. 算法简介
论文信息：
> [Real-time Scene Text Detection with Differentiable Binarization](https://arxiv.org/abs/1911.08947)
> Liao, Minghui and Wan, Zhaoyi and Yao, Cong and Chen, Kai and Bai, Xiang
> AAAI, 2020
在ICDAR2015文本检测公开数据集上，算法复现效果如下：
|模型|骨干网络|配置文件|precision|recall|Hmean|下载链接|
| --- | --- | --- | --- | --- | --- | --- |
|DB|ResNet50_vd|[configs/det/det_r50_vd_db.yml](../../configs/det/det_r50_vd_db.yml)|86.41%|78.72%|82.38%|[训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_db_v2.0_train.tar)|
|DB|MobileNetV3|[configs/det/det_mv3_db.yml](../../configs/det/det_mv3_db.yml)|77.29%|73.08%|75.12%|[训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_mv3_db_v2.0_train.tar)|
<a name="2"></a>
## 2. 环境配置
请先参考[《运行环境准备》](./environment.md)配置PaddleOCR运行环境，参考[《项目克隆》](./clone.md)克隆项目代码。
<a name="3"></a>
## 3. 模型训练、评估、预测
请参考[文本检测训练教程](./detection.md)。PaddleOCR对代码进行了模块化，训练不同的检测模型只需要**更换配置文件**即可。
<a name="4"></a>
## 4. 推理部署
<a name="4-1"></a>
### 4.1 Python推理
首先将DB文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_r50_vd_db_v2.0_train.tar) )，可以使用如下命令进行转换：
```shell
python3 tools/export_model.py -c configs/det/det_r50_vd_db.yml -o Global.pretrained_model=./det_r50_vd_db_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_db
```
DB文本检测模型推理，可以执行如下命令：
```shell
python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_db/"
```
可视化文本检测结果默认保存到`./inference_results`文件夹里面，结果文件的名称前缀为'det_res'。结果示例如下：
![](../imgs_results/det_res_img_10_db.jpg)
**注意**：由于ICDAR2015数据集只有1000张训练图像，且主要针对英文场景，所以上述模型对中文文本图像检测效果会比较差。
<a name="4-2"></a>
### 4.2 C++推理
准备好推理模型后，参考[cpp infer](../../deploy/cpp_infer/)教程进行操作即可。
<a name="4-3"></a>
### 4.3 Serving服务化部署
准备好推理模型后，参考[pdserving](../../deploy/pdserving/)教程进行Serving服务化部署，包括Python Serving和C++ Serving两种模式。
<a name="4-4"></a>
### 4.4 更多推理部署
DB模型还支持以下推理部署方式：
- Paddle2ONNX推理：准备好推理模型后，参考[paddle2onnx](../../deploy/paddle2onnx/)教程操作。
<a name="5"></a>
## 5. FAQ
## 引用
```bibtex
@inproceedings{liao2020real,
title={Real-time scene text detection with differentiable binarization},
author={Liao, Minghui and Wan, Zhaoyi and Yao, Cong and Chen, Kai and Bai, Xiang},
booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
volume={34},
number={07},
pages={11474--11481},
year={2020}
}
```
\ No newline at end of file
doc/doc_ch/algorithm_det_fcenet.md 0 → 100644
View file @ aa59fca5
# FCENet
- [1. 算法简介](#1)
- [2. 环境配置](#2)
- [3. 模型训练、评估、预测](#3)
- [3.1 训练](#3-1)
- [3.2 评估](#3-2)
- [3.3 预测](#3-3)
- [4. 推理部署](#4)
- [4.1 Python推理](#4-1)
- [4.2 C++推理](#4-2)
- [4.3 Serving服务化部署](#4-3)
- [4.4 更多推理部署](#4-4)
- [5. FAQ](#5)
<a name="1"></a>
## 1. 算法简介
论文信息：
> [Fourier Contour Embedding for Arbitrary-Shaped Text Detection](https://arxiv.org/abs/2104.10442)
> Yiqin Zhu and Jianyong Chen and Lingyu Liang and Zhanghui Kuang and Lianwen Jin and Wayne Zhang
> CVPR, 2021
在CTW1500文本检测公开数据集上，算法复现效果如下：
| 模型 |骨干网络|配置文件|precision|recall|Hmean|下载链接|
|-----| --- | --- | --- | --- | --- | --- |
| FCE | ResNet50_dcn | [configs/det/det_r50_vd_dcn_fce_ctw.yml](../../configs/det/det_r50_vd_dcn_fce_ctw.yml)| 88.39%|82.18%|85.27%|[训练模型](https://paddleocr.bj.bcebos.com/contribution/det_r50_dcn_fce_ctw_v2.0_train.tar)|
<a name="2"></a>
## 2. 环境配置
请先参考[《运行环境准备》](./environment.md)配置PaddleOCR运行环境，参考[《项目克隆》](./clone.md)克隆项目代码。
<a name="3"></a>
## 3. 模型训练、评估、预测
上述FCE模型使用CTW1500文本检测公开数据集训练得到，数据集下载可参考 [ocr_datasets](./dataset/ocr_datasets.md)。
数据下载完成后，请参考[文本检测训练教程](./detection.md)进行训练。PaddleOCR对代码进行了模块化，训练不同的检测模型只需要**更换配置文件**即可。
<a name="4"></a>
## 4. 推理部署
<a name="4-1"></a>
### 4.1 Python推理
首先将FCE文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd_dcn骨干网络，在CTW1500英文数据集训练的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/contribution/det_r50_dcn_fce_ctw_v2.0_train.tar) )，可以使用如下命令进行转换：
```shell
python3 tools/export_model.py -c configs/det/det_r50_vd_dcn_fce_ctw.yml -o Global.pretrained_model=./det_r50_dcn_fce_ctw_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_fce
```
FCE文本检测模型推理，执行非弯曲文本检测，可以执行如下命令：
```shell
python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_fce/" --det_algorithm="FCE" --det_fce_box_type=quad
```
可视化文本检测结果默认保存到`./inference_results`文件夹里面，结果文件的名称前缀为'det_res'。结果示例如下：
![](../imgs_results/det_res_img_10_fce.jpg)
如果想执行弯曲文本检测，可以执行如下命令：
```shell
python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img623.jpg" --det_model_dir="./inference/det_fce/" --det_algorithm="FCE" --det_fce_box_type=poly
```
可视化文本检测结果默认保存到`./inference_results`文件夹里面，结果文件的名称前缀为'det_res'。结果示例如下：
![](../imgs_results/det_res_img623_fce.jpg)
**注意**：由于CTW1500数据集只有1000张训练图像，且主要针对英文场景，所以上述模型对中文文本图像检测效果会比较差。
<a name="4-2"></a>
### 4.2 C++推理
由于后处理暂未使用CPP编写，FCE文本检测模型暂不支持CPP推理。
<a name="4-3"></a>
### 4.3 Serving服务化部署
暂未支持
<a name="4-4"></a>
### 4.4 更多推理部署
暂未支持
<a name="5"></a>
## 5. FAQ
## 引用
```bibtex
@InProceedings{zhu2021fourier,
title={Fourier Contour Embedding for Arbitrary-Shaped Text Detection},
author={Yiqin Zhu and Jianyong Chen and Lingyu Liang and Zhanghui Kuang and Lianwen Jin and Wayne Zhang},
year={2021},
booktitle = {CVPR}
}
```
doc/doc_ch/algorithm_det_psenet.md 0 → 100644
View file @ aa59fca5
# PSENet
- [1. 算法简介](#1)
- [2. 环境配置](#2)
- [3. 模型训练、评估、预测](#3)
- [3.1 训练](#3-1)
- [3.2 评估](#3-2)
- [3.3 预测](#3-3)
- [4. 推理部署](#4)
- [4.1 Python推理](#4-1)
- [4.2 C++推理](#4-2)
- [4.3 Serving服务化部署](#4-3)
- [4.4 更多推理部署](#4-4)
- [5. FAQ](#5)
<a name="1"></a>
## 1. 算法简介
论文信息：
> [Shape robust text detection with progressive scale expansion network](https://arxiv.org/abs/1903.12473)
> Wang, Wenhai and Xie, Enze and Li, Xiang and Hou, Wenbo and Lu, Tong and Yu, Gang and Shao, Shuai
> CVPR, 2019
在ICDAR2015文本检测公开数据集上，算法复现效果如下：
|模型|骨干网络|配置文件|precision|recall|Hmean|下载链接|
| --- | --- | --- | --- | --- | --- | --- |
|PSE| ResNet50_vd | [configs/det/det_r50_vd_pse.yml](../../configs/det/det_r50_vd_pse.yml)| 85.81% |79.53%|82.55%|[训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.1/en_det/det_r50_vd_pse_v2.0_train.tar)|
|PSE| MobileNetV3| [configs/det/det_mv3_pse.yml](../../configs/det/det_mv3_pse.yml) | 82.20% |70.48%|75.89%|[训练模型](https://paddleocr.bj.bcebos.com/dygraph_v2.1/en_det/det_mv3_pse_v2.0_train.tar)|
<a name="2"></a>
## 2. 环境配置
请先参考[《运行环境准备》](./environment.md)配置PaddleOCR运行环境，参考[《项目克隆》](./clone.md)克隆项目代码。
<a name="3"></a>
## 3. 模型训练、评估、预测
上述PSE模型使用ICDAR2015文本检测公开数据集训练得到，数据集下载可参考 [ocr_datasets](./dataset/ocr_datasets.md)。
数据下载完成后，请参考[文本检测训练教程](./detection.md)进行训练。PaddleOCR对代码进行了模块化，训练不同的检测模型只需要**更换配置文件**即可。
<a name="4"></a>
## 4. 推理部署
<a name="4-1"></a>
### 4.1 Python推理
首先将PSE文本检测训练过程中保存的模型，转换成inference model。以基于Resnet50_vd骨干网络，在ICDAR2015英文数据集训练的模型为例（ [模型下载地址](https://paddleocr.bj.bcebos.com/dygraph_v2.1/en_det/det_r50_vd_pse_v2.0_train.tar) )，可以使用如下命令进行转换：
```shell
python3 tools/export_model.py -c configs/det/det_r50_vd_pse.yml -o Global.pretrained_model=./det_r50_vd_pse_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_pse
```
PSE文本检测模型推理，执行非弯曲文本检测，可以执行如下命令：
```shell
python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_pse/" --det_algorithm="PSE" --det_pse_box_type=quad
```
可视化文本检测结果默认保存到`./inference_results`文件夹里面，结果文件的名称前缀为'det_res'。结果示例如下：
![](../imgs_results/det_res_img_10_pse.jpg)
如果想执行弯曲文本检测，可以执行如下命令：
```shell
python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_pse/" --det_algorithm="PSE" --det_pse_box_type=poly
```
可视化文本检测结果默认保存到`./inference_results`文件夹里面，结果文件的名称前缀为'det_res'。结果示例如下：
![](../imgs_results/det_res_img_10_pse_poly.jpg)
**注意**：由于ICDAR2015数据集只有1000张训练图像，且主要针对英文场景，所以上述模型对中文或弯曲文本图像检测效果会比较差。
<a name="4-2"></a>
### 4.2 C++推理
由于后处理暂未使用CPP编写，PSE文本检测模型暂不支持CPP推理。
<a name="4-3"></a>
### 4.3 Serving服务化部署
暂未支持
<a name="4-4"></a>
### 4.4 更多推理部署
暂未支持
<a name="5"></a>
## 5. FAQ
## 引用
```bibtex
@inproceedings{wang2019shape,
title={Shape robust text detection with progressive scale expansion network},
author={Wang, Wenhai and Xie, Enze and Li, Xiang and Hou, Wenbo and Lu, Tong and Yu, Gang and Shao, Shuai},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={9336--9345},
year={2019}
}
```
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment