Skip to content

Commit

Permalink
Merge pull request #6074 from MissPenguin/dygraph
Browse files Browse the repository at this point in the history 
update docs
  • Loading branch information
@MissPenguin
MissPenguin committed Apr 27, 2022
2 parents 4cddec7 + d417e22 commit 2b6c887
Show file tree
Hide file tree
Showing 5 changed files with 21 additions and 11 deletions.
6 changes: 4 additions & 2 deletions deploy/README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -22,9 +22,11 @@ PP-OCR has supported muti deployment schemes. Click the link to get the specific

- [Python Inference](../doc/doc_en/inference_ppocr_en.md)
- [C++ Inference](./cpp_infer/readme.md)
- [Serving](./pdserving/README.md)
- [Paddle-Lite](./lite/readme.md)
- [Serving (Python/C++)](./pdserving/README.md)
- [Paddle-Lite (ARM CPU/OpenCL ARM GPU/Metal ARM GPU)](./lite/readme.md)
- [Paddle.js](./paddlejs/README.md)
- [Jetson Inference]()
- [XPU Inference]()
- [Paddle2ONNX](./paddle2onnx/readme.md)

If you need the deployment tutorial of academic algorithm models other than PP-OCR, please directly enter the main page of corresponding algorithms, [entrance](../doc/doc_en/algorithm_overview_en.md)。
8 changes: 5 additions & 3 deletions deploy/README_ch.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -22,9 +22,11 @@ PP-OCR模型已打通多种场景部署方案，点击链接获取具体的使

- [Python 推理](../doc/doc_ch/inference_ppocr.md)
- [C++ 推理](./cpp_infer/readme_ch.md)
- [Serving 服务化部署](./pdserving/README_CN.md)
- [Paddle-Lite 端侧部署](./lite/readme_ch.md)
- [Paddle.js 服务化部署](./paddlejs/README_ch.md)
- [Serving 服务化部署（Python/C++）](./pdserving/README_CN.md)
- [Paddle-Lite 端侧部署（ARM CPU/OpenCL ARM GPU/Metal ARM GPU）](./lite/readme_ch.md)
- [Paddle.js 部署](./paddlejs/README_ch.md)
- [Jetson 推理]()
- [XPU 推理]()
- [Paddle2ONNX 推理](./paddle2onnx/readme_ch.md)

需要PP-OCR以外的学术算法模型的推理部署，请直接进入相应算法主页面，[入口](../doc/doc_ch/algorithm_overview.md)。
4 changes: 2 additions & 2 deletions doc/doc_ch/algorithm_det_db.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -25,8 +25,8 @@

|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|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|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)|
|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>
Expand Down
10 changes: 8 additions & 2 deletions doc/doc_ch/ppocr_introduction.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,8 @@

PP-OCR是PaddleOCR自研的实用的超轻量OCR系统。在实现[前沿算法](algorithm.md)的基础上，考虑精度与速度的平衡，进行**模型瘦身**和**深度优化**，使其尽可能满足产业落地需求。

#### PP-OCR

PP-OCR是一个两阶段的OCR系统，其中文本检测算法选用[DB](algorithm_det_db.md)，文本识别算法选用[CRNN](algorithm_rec_crnn.md)，并在检测和识别模块之间添加[文本方向分类器](angle_class.md)，以应对不同方向的文本识别。

PP-OCR系统pipeline如下：
Expand All @@ -28,9 +30,13 @@ PP-OCR系统pipeline如下：

PP-OCR系统在持续迭代优化，目前已发布PP-OCR和PP-OCRv2两个版本：

[1] PP-OCR从骨干网络选择和调整、预测头部的设计、数据增强、学习率变换策略、正则化参数选择、预训练模型使用以及模型自动裁剪量化8个方面，采用19个有效策略，对各个模块的模型进行效果调优和瘦身(如绿框所示)，最终得到整体大小为3.5M的超轻量中英文OCR和2.8M的英文数字OCR。更多细节请参考PP-OCR技术方案 https://arxiv.org/abs/2009.09941
PP-OCR从骨干网络选择和调整、预测头部的设计、数据增强、学习率变换策略、正则化参数选择、预训练模型使用以及模型自动裁剪量化8个方面，采用19个有效策略，对各个模块的模型进行效果调优和瘦身(如绿框所示)，最终得到整体大小为3.5M的超轻量中英文OCR和2.8M的英文数字OCR。更多细节请参考PP-OCR技术方案 https://arxiv.org/abs/2009.09941

#### PP-OCRv2

PP-OCRv2在PP-OCR的基础上，进一步在5个方面重点优化，检测模型采用CML协同互学习知识蒸馏策略和CopyPaste数据增广策略；识别模型采用LCNet轻量级骨干网络、UDML 改进知识蒸馏策略和[Enhanced CTC loss](./doc/doc_ch/enhanced_ctc_loss.md)损失函数改进（如上图红框所示），进一步在推理速度和预测效果上取得明显提升。更多细节请参考PP-OCRv2[技术报告](https://arxiv.org/abs/2109.03144)。

[2] PP-OCRv2在PP-OCR的基础上，进一步在5个方面重点优化，检测模型采用CML协同互学习知识蒸馏策略和CopyPaste数据增广策略；识别模型采用LCNet轻量级骨干网络、UDML 改进知识蒸馏策略和[Enhanced CTC loss](./doc/doc_ch/enhanced_ctc_loss.md)损失函数改进（如上图红框所示），进一步在推理速度和预测效果上取得明显提升。更多细节请参考PP-OCRv2[技术报告](https://arxiv.org/abs/2109.03144)。
#### PP-OCRv3


<a name="2"></a>
Expand Down
4 changes: 2 additions & 2 deletions doc/doc_en/algorithm_det_db_en.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -25,8 +25,8 @@ On the ICDAR2015 dataset, the text detection result is as follows:

|Model|Backbone|Configuration|Precision|Recall|Hmean|Download|
| --- | --- | --- | --- | --- | --- | --- |
|DB|ResNet50_vd|configs/det/det_r50_vd_db.yml|86.41%|78.72%|82.38%|[trained model](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|77.29%|73.08%|75.12%|[trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_mv3_db_v2.0_train.tar)|
|DB|ResNet50_vd|[configs/det/det_r50_vd_db.yml](../../configs/det/det_r50_vd_db.yml)|86.41%|78.72%|82.38%|[trained model](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%|[trained model](https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_mv3_db_v2.0_train.tar)|


<a name="2"></a>
Expand Down

0 comments on commit 2b6c887

Please sign in to comment.