The inference model (the model saved by paddle.jit.save) is generally a solidified model saved after the model training is completed, and is mostly used to give prediction in deployment.
The model saved during the training process is the checkpoints model, which saves the parameters of the model and is mostly used to resume training.
Compared with the checkpoints model, the inference model will additionally save the structural information of the model. Therefore, it is easier to deploy because the model structure and model parameters are already solidified in the inference model file, and is suitable for integration with actual systems. For more details, please refer to the document Classification Framework.
Next, we first introduce how to convert a trained model into an inference model, and then we will introduce text detection, text recognition, angle class, and the concatenation of them based on inference model.
- Inference Based on Python Prediction Engine
Download the lightweight Chinese detection model:
wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar && tar xf ./ch_lite/ch_ppocr_mobile_v2.0_det_train.tar -C ./ch_lite/
The above model is a DB algorithm trained with MobileNetV3 as the backbone. To convert the trained model into an inference model, just run the following command:
# -c Set the training algorithm yml configuration file
# -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_det_train/best_accuracy Global.save_inference_dir=./inference/det_db/
When converting to an inference model, the configuration file used is the same as the configuration file used during training. In addition, you also need to set the Global.pretrained_model parameter in the configuration file.
After the conversion is successful, there are three files in the model save directory:
inference/det_db/
├── inference.pdiparams # The parameter file of detection inference model
├── inference.pdiparams.info # The parameter information of detection inference model, which can be ignored
└── inference.pdmodel # The program file of detection inference model
Download the lightweight Chinese recognition model:
wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_train.tar && tar xf ./ch_lite/ch_ppocr_mobile_v2.0_rec_train.tar -C ./ch_lite/
The recognition model is converted to the inference model in the same way as the detection, as follows:
# -c Set the training algorithm yml configuration file
# -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/rec/ch_ppocr_v2.0/rec_chinese_lite_train_v2.0.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_rec_train/best_accuracy Global.save_inference_dir=./inference/rec_crnn/
If you have a model trained on your own dataset with a different dictionary file, please make sure that you modify the character_dict_path in the configuration file to your dictionary file path.
After the conversion is successful, there are three files in the model save directory:
inference/det_db/
├── inference.pdiparams # The parameter file of recognition inference model
├── inference.pdiparams.info # The parameter information of recognition inference model, which can be ignored
└── inference.pdmodel # The program file of recognition model
Download the angle classification model:
wget -P ./ch_lite/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_train.tar && tar xf ./ch_lite/ch_ppocr_mobile_v2.0_cls_train.tar -C ./ch_lite/
The angle classification model is converted to the inference model in the same way as the detection, as follows:
# -c Set the training algorithm yml configuration file
# -o Set optional parameters
# Global.pretrained_model parameter Set the training model address to be converted without adding the file suffix .pdmodel, .pdopt or .pdparams.
# Global.save_inference_dir Set the address where the converted model will be saved.
python3 tools/export_model.py -c configs/cls/cls_mv3.yml -o Global.pretrained_model=./ch_lite/ch_ppocr_mobile_v2.0_cls_train/best_accuracy Global.save_inference_dir=./inference/cls/
After the conversion is successful, there are two files in the directory:
inference/det_db/
├── inference.pdiparams # The parameter file of angle class inference model
├── inference.pdiparams.info # The parameter information of angle class inference model, which can be ignored
└── inference.pdmodel # The program file of angle class model
The following will introduce the lightweight Chinese detection model inference, DB text detection model inference and EAST text detection model inference. The default configuration is based on the inference setting of the DB text detection model. Because EAST and DB algorithms are very different, when inference, it is necessary to adapt the EAST text detection algorithm by passing in corresponding parameters.
For lightweight Chinese detection model inference, you can execute the following commands:
# download DB text detection inference model
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tar
tar xf ch_ppocr_mobile_v2.0_det_infer.tar
# predict
python3 tools/infer/predict_det.py --image_dir="./doc/imgs/00018069.jpg" --det_model_dir="./inference/det_db/"
The visual text detection results are saved to the ./inference_results folder by default, and the name of the result file is prefixed with 'det_res'. Examples of results are as follows:
You can use the parameters limit_type and det_limit_side_len to limit the size of the input image,
The optional parameters of limit_type are [max, min], and
det_limit_size_len is a positive integer, generally set to a multiple of 32, such as 960.
The default setting of the parameters is limit_type='max', det_limit_side_len=960. Indicates that the longest side of the network input image cannot exceed 960,
If this value is exceeded, the image will be resized with the same width ratio to ensure that the longest side is det_limit_side_len.
Set as limit_type='min', det_limit_side_len=960, it means that the shortest side of the image is limited to 960.
If the resolution of the input picture is relatively large and you want to use a larger resolution prediction, you can set det_limit_side_len to the desired value, such as 1216:
python3 tools/infer/predict_det.py --image_dir="./doc/imgs/1.jpg" --det_model_dir="./inference/det_db/" --det_limit_type=max --det_limit_side_len=1216
If you want to use the CPU for prediction, execute the command as follows
python3 tools/infer/predict_det.py --image_dir="./doc/imgs/1.jpg" --det_model_dir="./inference/det_db/" --use_gpu=False
First, convert the model saved in the DB text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example (model download link), you can use the following command to convert:
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 text detection model inference, you can execute the following command:
python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_db/"
The visualized text detection results are saved to the ./inference_results folder by default, and the name of the result file is prefixed with 'det_res'. Examples of results are as follows:
Note: Since the ICDAR2015 dataset has only 1,000 training images, mainly for English scenes, the above model has very poor detection result on Chinese text images.
First, convert the model saved in the EAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example (model download link), you can use the following command to convert:
python3 tools/export_model.py -c configs/det/det_r50_vd_east.yml -o Global.pretrained_model=./det_r50_vd_east_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_east
For EAST text detection model inference, you need to set the parameter --det_algorithm="EAST", run the following command:
python3 tools/infer/predict_det.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_east/" --det_algorithm="EAST"
The visualized text detection results are saved to the ./inference_results folder by default, and the name of the result file is prefixed with 'det_res'. Examples of results are as follows:
Note: EAST post-processing locality aware NMS has two versions: Python and C++. The speed of C++ version is obviously faster than that of Python version. Due to the compilation version problem of NMS of C++ version, C++ version NMS will be called only in Python 3.5 environment, and python version NMS will be called in other cases.
First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the ICDAR2015 English dataset as an example (model download link), you can use the following command to convert:
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_icdar15.yml -o Global.pretrained_model=./det_r50_vd_sast_icdar15_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_sast_ic15
For SAST quadrangle text detection model inference, you need to set the parameter --det_algorithm="SAST", run the following command:
python3 tools/infer/predict_det.py --det_algorithm="SAST" --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_sast_ic15/"
The visualized text detection results are saved to the ./inference_results folder by default, and the name of the result file is prefixed with 'det_res'. Examples of results are as follows:
First, convert the model saved in the SAST text detection training process into an inference model. Taking the model based on the Resnet50_vd backbone network and trained on the Total-Text English dataset as an example (model download link), you can use the following command to convert:
python3 tools/export_model.py -c configs/det/det_r50_vd_sast_totaltext.yml -o Global.pretrained_model=./det_r50_vd_sast_totaltext_v2.0_train/best_accuracy Global.save_inference_dir=./inference/det_sast_tt
For SAST curved text detection model inference, you need to set the parameter --det_algorithm="SAST" and --det_sast_polygon=True, run the following command:
python3 tools/infer/predict_det.py --det_algorithm="SAST" --image_dir="./doc/imgs_en/img623.jpg" --det_model_dir="./inference/det_sast_tt/" --det_sast_polygon=True
The visualized text detection results are saved to the ./inference_results folder by default, and the name of the result file is prefixed with 'det_res'. Examples of results are as follows:
Note: SAST post-processing locality aware NMS has two versions: Python and C++. The speed of C++ version is obviously faster than that of Python version. Due to the compilation version problem of NMS of C++ version, C++ version NMS will be called only in Python 3.5 environment, and python version NMS will be called in other cases.
The following will introduce the lightweight Chinese recognition model inference, other CTC-based and Attention-based text recognition models inference. For Chinese text recognition, it is recommended to choose the recognition model based on CTC loss. In practice, it is also found that the result of the model based on Attention loss is not as good as the one based on CTC loss. In addition, if the characters dictionary is modified during training, make sure that you use the same characters set during inference. Please check below for details.
For lightweight Chinese recognition model inference, you can execute the following commands:
# download CRNN text recognition inference model
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tar
tar xf ch_ppocr_mobile_v2.0_rec_infer.tar
python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words_en/word_10.png" --rec_model_dir="ch_ppocr_mobile_v2.0_rec_infer"
After executing the command, the prediction results (recognized text and score) of the above image will be printed on the screen.
Predicts of ./doc/imgs_words_en/word_10.png:('PAIN', 0.9897658)Taking CRNN as an example, we introduce the recognition model inference based on CTC loss. Rosetta and Star-Net are used in a similar way, No need to set the recognition algorithm parameter rec_algorithm.
First, convert the model saved in the CRNN text recognition training process into an inference model. Taking the model based on Resnet34_vd backbone network, using MJSynth and SynthText (two English text recognition synthetic datasets) for training, as an example (model download address). It can be converted as follow:
python3 tools/export_model.py -c configs/det/rec_r34_vd_none_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_none_bilstm_ctc_v2.0_train/best_accuracy Global.save_inference_dir=./inference/rec_crnn
For CRNN text recognition model inference, execute the following commands:
python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words_en/word_336.png" --rec_model_dir="./inference/starnet/" --rec_image_shape="3, 32, 100" --rec_char_dict_path="./ppocr/utils/ic15_dict.txt"
After executing the command, the recognition result of the above image is as follows:
Predicts of ./doc/imgs_words_en/word_336.png:('super', 0.9999073)Note：Since the above model refers to DTRB text recognition training and evaluation process, it is different from the training of lightweight Chinese recognition model in two aspects:
-
The image resolution used in training is different: the image resolution used in training the above model is [3，32，100], while during our Chinese model training, in order to ensure the recognition effect of long text, the image resolution used in training is [3, 32, 320]. The default shape parameter of the inference stage is the image resolution used in training phase, that is [3, 32, 320]. Therefore, when running inference of the above English model here, you need to set the shape of the recognition image through the parameter
rec_image_shape. -
Character list: the experiment in the DTRB paper is only for 26 lowercase English characters and 10 numbers, a total of 36 characters. All upper and lower case characters are converted to lower case characters, and characters not in the above list are ignored and considered as spaces. Therefore, no characters dictionary file is used here, but a dictionary is generated by the below command.
self.character_str = "0123456789abcdefghijklmnopqrstuvwxyz"
dict_character = list(self.character_str)
The recognition model based on SRN requires additional setting of the recognition algorithm parameter --rec_algorithm="SRN". At the same time, it is necessary to ensure that the predicted shape is consistent with the training, such as: --rec_image_shape="1, 64, 256"
python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words_en/word_336.png" \
--rec_model_dir="./inference/srn/" \
--rec_image_shape="1, 64, 256" \
--rec_char_dict_path="./ppocr/utils/ic15_dict.txt" \
--rec_algorithm="SRN"
If the text dictionary is modified during training, when using the inference model to predict, you need to specify the dictionary path used by --rec_char_dict_path
python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words_en/word_336.png" --rec_model_dir="./your inference model" --rec_image_shape="3, 32, 100" --rec_char_dict_path="your text dict path"
If you need to predict other language models, when using inference model prediction, you need to specify the dictionary path used by --rec_char_dict_path. At the same time, in order to get the correct visualization results,
You need to specify the visual font path through --vis_font_path. There are small language fonts provided by default under the doc/fonts path, such as Korean recognition:
python3 tools/infer/predict_rec.py --image_dir="./doc/imgs_words/korean/1.jpg" --rec_model_dir="./your inference model" --rec_char_dict_path="ppocr/utils/dict/korean_dict.txt" --vis_font_path="doc/fonts/korean.ttf"
After executing the command, the prediction result of the above figure is:
Predicts of ./doc/imgs_words/korean/1.jpg:('바탕으로', 0.9948904)
For angle classification model inference, you can execute the following commands:
python3 tools/infer/predict_cls.py --image_dir="./doc/imgs_words_en/word_10.png" --cls_model_dir="./inference/cls/"
# download text angle class inference model：
wget https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar
tar xf ch_ppocr_mobile_v2.0_cls_infer.tar
python3 tools/infer/predict_cls.py --image_dir="./doc/imgs_words_en/word_10.png" --cls_model_dir="ch_ppocr_mobile_v2.0_cls_infer"
After executing the command, the prediction results (classification angle and score) of the above image will be printed on the screen.
Predicts of ./doc/imgs_words_en/word_10.png:['0', 0.9999995]
When performing prediction, you need to specify the path of a single image or a folder of images through the parameter image_dir, the parameter det_model_dir specifies the path to detect the inference model, the parameter cls_model_dir specifies the path to angle classification inference model and the parameter rec_model_dir specifies the path to identify the inference model. The parameter use_angle_cls is used to control whether to enable the angle classification model. The parameter use_mp specifies whether to use multi-process to infer total_process_num specifies process number when using multi-process. The parameter(Paddle Inference is not thread-safe, it is recommended to use multi-process) . The visualized recognition results are saved to the ./inference_results folder by default.
# use direction classifier
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/00018069.jpg" --det_model_dir="./inference/det_db/" --cls_model_dir="./inference/cls/" --rec_model_dir="./inference/rec_crnn/" --use_angle_cls=true
# not use use direction classifier
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/00018069.jpg" --det_model_dir="./inference/det_db/" --rec_model_dir="./inference/rec_crnn/"
# use multi-process
python3 tools/infer/predict_system.py --image_dir="./doc/imgs/00018069.jpg" --det_model_dir="./inference/det_db/" --rec_model_dir="./inference/rec_crnn/" --use_angle_cls=false --use_mp=True --total_process_num=6After executing the command, the recognition result image is as follows:
If you want to try other detection algorithms or recognition algorithms, please refer to the above text detection model inference and text recognition model inference, update the corresponding configuration and model.
Note: due to the limitation of rotation logic of detected box, SAST curved text detection model (using the parameter det_sast_polygon=True) is not supported for model combination yet.
The following command uses the combination of the EAST text detection and STAR-Net text recognition:
python3 tools/infer/predict_system.py --image_dir="./doc/imgs_en/img_10.jpg" --det_model_dir="./inference/det_east/" --det_algorithm="EAST" --rec_model_dir="./inference/starnet/" --rec_image_shape="3, 32, 100" --rec_char_dict_path="./ppocr/utils/ic15_dict.txt"
After executing the command, the recognition result image is as follows:
