English | 简体中文

Introduction

PaddleOCR aims to create a rich, leading, and practical OCR tools that help users train better models and apply them into practice.

Recent updates

• 2020.6.8 Add dataset and keep updating
• 2020.6.5 Support exporting attention model to inference_model
• 2020.6.5 Support separate prediction and recognition, output result score
• 2020.5.30 Provide ultra-lightweight Chinese OCR online experience
• 2020.5.30 Model prediction and training supported on Windows system
• more

Features

• Ultra-lightweight Chinese OCR model, total model size is only 8.6M
  • Single model supports Chinese and English numbers combination recognition, vertical text recognition, long text recognition
  • Detection model DB (4.1M) + recognition model CRNN (4.5M)
• Various text detection algorithms: EAST, DB
• Various text recognition algorithms: Rosetta, CRNN, STAR-Net, RARE

Supported Chinese models list:

Model Name	Description	Detection Model link	Recognition Model link
chinese_db_crnn_mobile	Ultra-lightweight Chinese OCR model	inference model & pre-trained model	inference model & pre-trained model
chinese_db_crnn_server	General Chinese OCR model	inference model & pre-trained model	inference model & pre-trained model

For testing our Chinese OCR online：https://www.paddlepaddle.org.cn/hub/scene/ocr

You can also quickly experience the Ultra-lightweight Chinese OCR and General Chinese OCR models as follows:

Ultra-lightweight Chinese OCR and General Chinese OCR inference

The picture above is the result of our Ultra-lightweight Chinese OCR model. For more testing results, please see the end of the article Ultra-lightweight Chinese OCR results and General Chinese OCR results.

1. Environment configuration

Please see Quick installation

2. Download inference models

(1) Download Ultra-lightweight Chinese OCR models

If wget is not installed in the windows system, you can copy the link to the browser to download the model. After model downloaded, unzip it and place it in the corresponding directory

mkdir inference && cd inference
# Download the detection part of the Ultra-lightweight Chinese OCR and decompress it
wget https://paddleocr.bj.bcebos.com/ch_models/ch_det_mv3_db_infer.tar && tar xf ch_det_mv3_db_infer.tar
# Download the recognition part of the Ultra-lightweight Chinese OCR and decompress it
wget https://paddleocr.bj.bcebos.com/ch_models/ch_rec_mv3_crnn_infer.tar && tar xf ch_rec_mv3_crnn_infer.tar
cd ..

(2) Download General Chinese OCR models

mkdir inference && cd inference
# Download the detection part of the general Chinese OCR model and decompress it
wget https://paddleocr.bj.bcebos.com/ch_models/ch_det_r50_vd_db_infer.tar && tar xf ch_det_r50_vd_db_infer.tar
# Download the recognition part of the generic Chinese OCR model and decompress it
wget https://paddleocr.bj.bcebos.com/ch_models/ch_rec_r34_vd_crnn_infer.tar && tar xf ch_rec_r34_vd_crnn_infer.tar
cd ..

3. Single image and batch image prediction

The following code implements text detection and recognition inference tandemly. When performing prediction, you need to specify the path of a single image or image folder through the parameter image_dir, the parameter det_model_dir specifies the path to detection model, and the parameter rec_model_dir specifies the path to the recognition model. The visual prediction results are saved to the ./inference_results folder by default.

# Prediction on a single image by specifying image path to image_dir
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/11.jpg" --det_model_dir="./inference/ch_det_mv3_db/"  --rec_model_dir="./inference/ch_rec_mv3_crnn/"

# Prediction on a batch of images by specifying image folder path to image_dir
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/" --det_model_dir="./inference/ch_det_mv3_db/"  --rec_model_dir="./inference/ch_rec_mv3_crnn/"

# If you want to use CPU for prediction, you need to set the use_gpu parameter to False
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/11.jpg" --det_model_dir="./inference/ch_det_mv3_db/"  --rec_model_dir="./inference/ch_rec_mv3_crnn/" --use_gpu=False

To run inference of the Generic Chinese OCR model, follow these steps above to download the corresponding models and update the relevant parameters. Examples are as follows:

# Prediction on a single image by specifying image path to image_dir
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/11.jpg" --det_model_dir="./inference/ch_det_r50_vd_db/"  --rec_model_dir="./inference/ch_rec_r34_vd_crnn/"

For more text detection and recognition models, please refer to the document Inference

Documentation

• Quick installation
• Text detection model training/evaluation/prediction
• Text recognition model training/evaluation/prediction
• Inference
• Dataset

Text detection algorithm

PaddleOCR open source text detection algorithms list:

• EAST(paper)
• DB(paper)
• SAST(paper)(Baidu Self-Research, comming soon)

On the ICDAR2015 dataset, the text detection result is as follows:

Model	Backbone	precision	recall	Hmean	Download link
EAST	ResNet50_vd	88.18%	85.51%	86.82%	Download link
EAST	MobileNetV3	81.67%	79.83%	80.74%	Download link
DB	ResNet50_vd	83.79%	80.65%	82.19%	Download link
DB	MobileNetV3	75.92%	73.18%	74.53%	Download link

For use of LSVT street view dataset with a total of 3w training data，the related configuration and pre-trained models for Chinese detection task are as follows:

Model	Backbone	Configuration file	Pre-trained model
Ultra-lightweight Chinese model	MobileNetV3	det_mv3_db.yml	Download link
General Chinese OCR model	ResNet50_vd	det_r50_vd_db.yml	Download link

• Note: For the training and evaluation of the above DB model, post-processing parameters box_thresh=0.6 and unclip_ratio=1.5 need to be set. If using different datasets and different models for training, these two parameters can be adjusted for better result.

For the training guide and use of PaddleOCR text detection algorithms, please refer to the document Text detection model training/evaluation/prediction

Text recognition algorithm

PaddleOCR open-source text recognition algorithms list:

• CRNN(paper)
• Rosetta(paper)
• STAR-Net(paper)
• RARE(paper)
• SRN(paper)(Baidu Self-Research, comming soon)

Refer to DTRB, the training and evaluation result of these above text recognition (using MJSynth and SynthText for training, evaluate on IIIT, SVT, IC03, IC13, IC15, SVTP, CUTE) is as follow:

Model	Backbone	Avg Accuracy	Module combination	Download link
Rosetta	Resnet34_vd	80.24%	rec_r34_vd_none_none_ctc	Download link
Rosetta	MobileNetV3	78.16%	rec_mv3_none_none_ctc	Download link
CRNN	Resnet34_vd	82.20%	rec_r34_vd_none_bilstm_ctc	Download link
CRNN	MobileNetV3	79.37%	rec_mv3_none_bilstm_ctc	Download link
STAR-Net	Resnet34_vd	83.93%	rec_r34_vd_tps_bilstm_ctc	Download link
STAR-Net	MobileNetV3	81.56%	rec_mv3_tps_bilstm_ctc	Download link
RARE	Resnet34_vd	84.90%	rec_r34_vd_tps_bilstm_attn	Download link
RARE	MobileNetV3	83.32%	rec_mv3_tps_bilstm_attn	Download link

We use LSVT dataset and cropout 30w traning data from original photos by using position groundtruth and make some calibration needed. In addition, based on the LSVT corpus, 500w synthetic data is generated to train the Chinese model. The related configuration and pre-trained models are as follows:

Model	Backbone	Configuration file	Pre-trained model
Ultra-lightweight Chinese model	MobileNetV3	rec_chinese_lite_train.yml	Download link
General Chinese OCR model	Resnet34_vd	rec_chinese_common_train.yml	Download link

Please refer to the document for training guide and use of PaddleOCR text recognition algorithms Text recognition model training/evaluation/prediction

End-to-end OCR algorithm

• End2End-PSL(Baidu Self-Research, comming soon)

Ultra-lightweight Chinese OCR results

General Chinese OCR results

FAQ

1. Error when using attention-based recognition model: KeyError: 'predict'

   The inference of recognition model based on attention loss is still being debugged. For Chinese text recognition, it is recommended to choose the recognition model based on CTC loss first. In practice, it is also found that the recognition model based on attention loss is not as effective as the one based on CTC loss.

2. About inference speed

   When there are a lot of texts in the picture, the prediction time will increase. You can use --rec_batch_num to set a smaller prediction batch size. The default value is 30, which can be changed to 10 or other values.

3. Service deployment and mobile deployment

   It is expected that the service deployment based on Serving and the mobile deployment based on Paddle Lite will be released successively in mid-to-late June. Stay tuned for more updates.

4. Release time of self-developed algorithm

   Baidu Self-developed algorithms such as SAST, SRN and end2end PSL will be released in June or July. Please be patient.

more

Welcome to the PaddleOCR technical exchange group

WeChat: paddlehelp . remarks OCR, the assistant will invite you to join the group~

References

1. EAST:
@inproceedings{zhou2017east,
  title={EAST: an efficient and accurate scene text detector},
  author={Zhou, Xinyu and Yao, Cong and Wen, He and Wang, Yuzhi and Zhou, Shuchang and He, Weiran and Liang, Jiajun},
  booktitle={Proceedings of the IEEE conference on Computer Vision and Pattern Recognition},
  pages={5551--5560},
  year={2017}
}

2. DB:
@article{liao2019real,
  title={Real-time Scene Text Detection with Differentiable Binarization},
  author={Liao, Minghui and Wan, Zhaoyi and Yao, Cong and Chen, Kai and Bai, Xiang},
  journal={arXiv preprint arXiv:1911.08947},
  year={2019}
}

3. DTRB:
@inproceedings{baek2019wrong,
  title={What is wrong with scene text recognition model comparisons? dataset and model analysis},
  author={Baek, Jeonghun and Kim, Geewook and Lee, Junyeop and Park, Sungrae and Han, Dongyoon and Yun, Sangdoo and Oh, Seong Joon and Lee, Hwalsuk},
  booktitle={Proceedings of the IEEE International Conference on Computer Vision},
  pages={4715--4723},
  year={2019}
}

4. SAST:
@inproceedings{wang2019single,
  title={A Single-Shot Arbitrarily-Shaped Text Detector based on Context Attended Multi-Task Learning},
  author={Wang, Pengfei and Zhang, Chengquan and Qi, Fei and Huang, Zuming and En, Mengyi and Han, Junyu and Liu, Jingtuo and Ding, Errui and Shi, Guangming},
  booktitle={Proceedings of the 27th ACM International Conference on Multimedia},
  pages={1277--1285},
  year={2019}
}

5. SRN:
@article{yu2020towards,
  title={Towards Accurate Scene Text Recognition with Semantic Reasoning Networks},
  author={Yu, Deli and Li, Xuan and Zhang, Chengquan and Han, Junyu and Liu, Jingtuo and Ding, Errui},
  journal={arXiv preprint arXiv:2003.12294},
  year={2020}
}

6. end2end-psl:
@inproceedings{sun2019chinese,
  title={Chinese Street View Text: Large-scale Chinese Text Reading with Partially Supervised Learning},
  author={Sun, Yipeng and Liu, Jiaming and Liu, Wei and Han, Junyu and Ding, Errui and Liu, Jingtuo},
  booktitle={Proceedings of the IEEE International Conference on Computer Vision},
  pages={9086--9095},
  year={2019}
}

License

This project is released under Apache 2.0 license

Contribution

We welcome all the contributions to PaddleOCR and appreciate for your feedback very much.

• Many thanks to Khanh Tran for contributing the English documentation.