Table Structure Recognition Module Development Tutorial¶

I. Overview¶

Table structure recognition is a crucial component in table recognition systems, converting non-editable table images into editable table formats (e.g., HTML). The goal of table structure recognition is to identify the rows, columns, and cell positions of tables. The performance of this module directly impacts the accuracy and efficiency of the entire table recognition system. The module typically outputs HTML or LaTeX code for the table area, which is then passed to the table content recognition module for further processing.

II. Supported Model List¶

Model	Model Download Link	Accuracy (%)	GPU Inference Time (ms) [Normal Mode / High-Performance Mode]	CPU Inference Time (ms) [Normal Mode / High-Performance Mode]	Model Size (M)	Description
SLANet	Inference Model/Trained Model	59.52	103.08 / 103.08	197.99 / 197.99	6.9 M	SLANet is a table structure recognition model developed by Baidu PaddlePaddle Vision Team. The model significantly improves the accuracy and inference speed of table structure recognition by adopting a CPU-friendly lightweight backbone network PP-LCNet, a high-low-level feature fusion module CSP-PAN, and a feature decoding module SLA Head that aligns structural and positional information.
SLANet_plus	Inference Model/Trained Model	63.69	140.29 / 140.29	195.39 / 195.39	6.9 M	SLANet_plus is an enhanced version of SLANet, a table structure recognition model developed by Baidu PaddlePaddle's Vision Team. Compared to SLANet, SLANet_plus significantly improves its recognition capabilities for wireless and complex tables, while reducing the model's sensitivity to the accuracy of table localization. Even when there are offsets in table localization, it can still perform relatively accurate recognition.
SLANeXt_wired	Inference Model/Trained Model	69.65			351M	SLANeXt series is a new generation of form structure recognition model developed by Baidu PaddlePaddle's Vision Team. Compared with SLANet and SLANet_plus, SLANeXt focuses on the recognition of form structure, and special weights are trained for the recognition of wired and wireless forms, which significantly improves the recognition ability of all types of forms, especially the recognition ability of wired forms is greatly improved.
SLANeXt_wireless	Inference Model/Trained Model	69.65			351M

Test Environment Description:

Performance Test Environment
Test Dataset: PaddleX Internal Self-built High-difficulty Chinese Table Recognition Dataset.
Hardware Configuration:
- GPU: NVIDIA Tesla T4
- CPU: Intel Xeon Gold 6271C @ 2.60GHz
- Other Environments: Ubuntu 20.04 / cuDNN 8.6 / TensorRT 8.5.2.2
Inference Mode Description

Mode	GPU Configuration	CPU Configuration	Acceleration Technology Combination
Normal Mode	FP32 Precision / No TRT Acceleration	FP32 Precision / 8 Threads	PaddleInference
High-Performance Mode	Optimal combination of pre-selected precision types and acceleration strategies	FP32 Precision / 8 Threads	Pre-selected optimal backend (Paddle/OpenVINO/TRT, etc.)

III. Quick Integration¶

❗ Before quick integration, please install the PaddleX wheel package. For detailed instructions, refer to PaddleX Local Installation Guide

After installing the wheel package, a few lines of code can complete the inference of the table structure recognition module. You can easily switch models within this module and integrate the model inference into your project. Before running the following code, please download the demo image to your local machine.

from paddlex import create_model
model = create_model(model_name="SLANet")
output = model.predict(input="table_recognition.jpg", batch_size=1)
for res in output:
    res.print(json_format=False)
    res.save_to_json("./output/res.json")

👉 After running, the result is: (Click to expand)

{'res': {'input_path': 'table_recognition.jpg', 'page_index': None, 'bbox': [array([ 42,   2, 390,   2, 388,  27,  40,  26]), array([11, 35, 89, 35, 87, 63, 11, 63]), array([113,  34, 192,  34, 186,  64, 109,  64]), array([219,  33, 399,  33, 393,  62, 212,  62]), array([413,  33, 544,  33, 544,  64, 407,  64]), array([12, 67, 98, 68, 96, 93, 12, 93]), array([115,  66, 205,  66, 200,  91, 111,  91]), array([234,  65, 390,  65, 385,  92, 227,  92]), array([414,  66, 537,  67, 537,  95, 409,  95]), array([  7,  97, 106,  97, 104, 128,   7, 128]), array([113,  96, 206,  95, 201, 127, 109, 127]), array([236,  96, 386,  96, 381, 128, 230, 128]), array([413,  96, 534,  95, 533, 127, 408, 127])], 'structure': ['<html>', '<body>', '<table>', '<tr>', '<td', ' colspan="4"', '>', '</td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '<tr>', '<td></td>', '<td></td>', '<td></td>', '<td></td>', '</tr>', '</table>', '</body>', '</html>'], 'structure_score': 0.99948007}}

Parameter meanings are as follows:

input_path: The path of the input image to be predicted
page_index：If the input is a PDF file, this indicates the current page number of the PDF. Otherwise, it is `None`
boxes: Predicted table cell information, a list composed of several predicted table cell coordinates. Note that the table cell predictions from the SLANeXt series models are invalid
structure: Predicted table structure in HTML expressions, a list composed of several predicted HTML keywords in order
structure_score: Confidence score of the predicted table structure

Relevant methods, parameters, and explanations are as follows:

create_model instantiates a table structure recognition model (here, SLANet is used as an example), with specific details as follows:

Parameter	Parameter Description	Parameter Type	Options	Default Value
`model_name`	Name of the model	`str`	All model names supported by PaddleX	None
`model_dir`	Path to store the model	`str`	None	None

model_name must be specified. After specifying model_name, the default model parameters from PaddleX will be used. If model_dir is specified, the user-defined model will be used.
The predict() method of the table structure recognition model is called for inference and prediction. The predict() method has parameters input and batch_size, with specific details as follows:

Parameter	Parameter Description	Parameter Type	Options	Default Value
`input`	Data to be predicted, supporting multiple input types	`Python Var`/`str`/`list`	Python Variable, such as image data represented by `numpy.ndarray` File Path, such as the local path of an image file: `/root/data/img.jpg` URL Link, such as the web URL of an image file: Example Local Directory, the directory should contain the data files to be predicted, such as the local path: `/root/data/` List, the elements of the list should be of the above-mentioned data types, such as `[numpy.ndarray, numpy.ndarray]`, `[\"/root/data/img1.jpg\", \"/root/data/img2.jpg\"]`, `[\"/root/data1\", \"/root/data2\"]`	None
`batch_size`	Batch size	`int`	Any integer	1

Process the prediction results. Each sample's prediction result is a corresponding Result object, and it supports operations such as printing and saving as a json file:

Method	Method Description	Parameter	Parameter Type	Parameter Description	Default Value
`print()`	Print the result to the terminal	`format_json`	`bool`	Whether to format the output content using `JSON` indentation	`True`
		`indent`	`int`	Specify the indentation level to beautify the output `JSON` data, making it more readable. It is only effective when `format_json` is `True`	4
		`ensure_ascii`	`bool`	Control whether to escape non-`ASCII` characters to `Unicode`. If set to `True`, all non-`ASCII` characters will be escaped; `False` will retain the original characters. It is only effective when `format_json` is `True`	`False`
`save_to_json()`	Save the result as a JSON-formatted file	`save_path`	`str`	The path to save the file. If it is a directory, the saved file will have the same name as the input file	None
		`indent`	`int`	Specify the indentation level to beautify the output `JSON` data, making it more readable. It is only effective when `format_json` is `True`	4
		`ensure_ascii`	`bool`	Control whether to escape non-`ASCII` characters to `Unicode`. If set to `True`, all non-`ASCII` characters will be escaped; `False` will retain the original characters. It is only effective when `format_json` is `True`	`False`

In addition, it also supports obtaining results through attributes, as follows:

Attribute	Attribute Description
`json`	Get the prediction result in `json` format

For more information on using PaddleX's single-model inference APIs, refer to PaddleX Single Model Python Script Usage Instructions.

IV. Custom Development¶

If you seek higher accuracy from existing models, you can leverage PaddleX's custom development capabilities to develop better table structure recognition models. Before developing table structure recognition models with PaddleX, ensure you have installed the PaddleOCR plugin for PaddleX. The installation process can be found in the PaddleX Local Installation Guide

4.1 Data Preparation¶

Before model training, you need to prepare the corresponding dataset for the task module. PaddleX provides data validation functionality for each module, and only data that passes validation can be used for model training. Additionally, PaddleX provides demo datasets for each module, which you can use to complete subsequent development. If you wish to use a private dataset for model training, refer to PaddleX Table Structure Recognition Task Module Data Annotation Tutorial

4.1.1 Demo Data Download¶

You can download the demo dataset to a specified folder using the following command:

wget https://paddle-model-ecology.bj.bcebos.com/paddlex/data/table_rec_dataset_examples.tar -P ./dataset
tar -xf ./dataset/table_rec_dataset_examples.tar -C ./dataset/

4.1.2 Data Validation¶

Run a single command to complete data validation:

python main.py -c paddlex/configs/modules/table_recognition/SLANet.yaml \
    -o Global.mode=check_dataset \
    -o Global.dataset_dir=./dataset/table_rec_dataset_examples

After executing the above command, PaddleX will validate the dataset and summarize its basic information. If the command runs successfully, it will print Check dataset passed ! in the log. The validation results file is saved in ./output/check_dataset_result.json, and related outputs are saved in the ./output/check_dataset directory in the current directory, including visual examples of sample images and sample distribution histograms.

👉 Details of Validation Results (Click to Expand)

The specific content of the validation result file is:

{
  "done_flag": true,
  "check_pass": true,
  "attributes": {
    "train_samples": 2000,
    "train_sample_paths": [
      "check_dataset/demo_img/border_right_7384_X9UFEPKVMLALY7DDB11A.jpg",
      "check_dataset/demo_img/border_top_13708_VE2DGBD4DCQU2ITLBTEA.jpg",
      "check_dataset/demo_img/border_top_6490_14Z6ZN6G52GG4XA0K4XU.jpg",
      "check_dataset/demo_img/border_top_14236_DG96EX0EDKIIDK8P6ENG.jpg",
      "check_dataset/demo_img/border_19648_SV8B7X34RTYRAT2T5CPI.jpg",
    ],
    "val_samples": 100,
    "val_sample_paths": [
      "check_dataset/demo_img/border_18653_J9SSKHLFTRJD4J8W17OW.jpg",
      "check_dataset/demo_img/border_bottom_8396_VJ3QJ3I0DP63P4JR77FE.jpg",
      "check_dataset/demo_img/border_9017_K2V7QBWSU2BA4R3AJSO7.jpg",
      "check_dataset/demo_img/border_top_19494_SDFMWP92NOB2OT7109FI.jpg",
      "check_dataset/demo_img/no_border_288_6LK683JUCMOQ38V5BV29.jpg"
    ]
  },
  "analysis": {},
  "dataset_path": "./dataset/table_rec_dataset_examples",
  "show_type": "image",
  "dataset_type": "PubTabTableRecDataset"
}

In the above validation results, check_pass being True indicates that the dataset format meets the requirements. Explanations for other indicators are as follows:

attributes.train_samples: The number of samples in the training set of this dataset is 2000;
attributes.val_samples: The number of samples in the validation set of this dataset is 100;
attributes.train_sample_paths: A list of relative paths to the visualization images of samples in the training set of this dataset;
attributes.val_sample_paths: A list of relative paths to the visualization images of samples in the validation set of this dataset.

4.1.3 Dataset Format Conversion / Dataset Splitting (Optional)¶

After completing the dataset verification, you can convert the dataset format or re-split the training/validation ratio by modifying the configuration file or appending hyperparameters.

👉 Details on Format Conversion / Dataset Splitting (Click to Expand)

(1) Dataset Format Conversion

Table structure recognition does not support data format conversion.

(2) Dataset Splitting

The dataset splitting parameters can be set by modifying the fields under CheckDataset in the configuration file. An example of part of the configuration file is shown below:

CheckDataset:
split:
enable: Whether to re-split the dataset. Set to True to enable dataset splitting, default is False;
train_percent: If re-splitting the dataset, set the percentage of the training set. The type is any integer between 0-100, ensuring the sum with val_percent equals 100;

For example, if you want to re-split the dataset with a 90% training set and a 10% validation set, modify the configuration file as follows:

......
CheckDataset:
  ......
  split:
    enable: True
    train_percent: 90
    val_percent: 10
  ......

Then execute the command:

python main.py -c paddlex/configs/modules/table_recognition/SLANet.yaml \
    -o Global.mode=check_dataset \
    -o Global.dataset_dir=./dataset/table_rec_dataset_examples

After the data splitting is executed, the original annotation files will be renamed to xxx.bak in their original paths.

The above parameters also support setting through appending command line arguments:

python main.py -c paddlex/configs/modules/table_recognition/SLANet.yaml \
    -o Global.mode=check_dataset \
    -o Global.dataset_dir=./dataset/table_rec_dataset_examples \
    -o CheckDataset.split.enable=True \
    -o CheckDataset.split.train_percent=90 \
    -o CheckDataset.split.val_percent=10

4.2 Model Training¶

A single command can complete the model training. Taking the training of the table structure recognition model SLANet as an example:

python main.py -c paddlex/configs/modules/table_recognition/SLANet.yaml \
    -o Global.mode=train \
    -o Global.dataset_dir=./dataset/table_rec_dataset_examples

the following steps are required:

Specify the path of the model's .yaml configuration file (here it is SLANet.yaml,When training other models, you need to specify the corresponding configuration files. The relationship between the model and configuration files can be found in the PaddleX Model List (CPU/GPU))
Specify the mode as model training: -o Global.mode=train
Specify the path of the training dataset: -o Global.dataset_dir. Other related parameters can be set by modifying the fields under Global and Train in the .yaml configuration file, or adjusted by appending parameters in the command line. For example, to specify training on the first 2 GPUs: -o Global.device=gpu:0,1; to set the number of training epochs to 10: -o Train.epochs_iters=10. For more modifiable parameters and their detailed explanations, refer to the configuration file parameter instructions for the corresponding task module of the model PaddleX Common Model Configuration File Parameters.

👉 More Details (Click to Expand)

During model training, PaddleX automatically saves the model weight files, with the default being output. If you need to specify a save path, you can set it through the -o Global.output field in the configuration file.
PaddleX shields you from the concepts of dynamic graph weights and static graph weights. During model training, both dynamic and static graph weights are produced, and static graph weights are selected by default for model inference.
After completing the model training, all outputs are saved in the specified output directory (default is ./output/), typically including:
train_result.json: Training result record file, recording whether the training task was completed normally, as well as the output weight metrics, related file paths, etc.;
train.log: Training log file, recording changes in model metrics and loss during training;
config.yaml: Training configuration file, recording the hyperparameter configuration for this training session;
.pdparams, .pdema, .pdopt.pdstate, .pdiparams, .pdmodel: Model weight-related files, including network parameters, optimizer, EMA, static graph network parameters, static graph network structure, etc.;

4.3 Model Evaluation¶

After completing model training, you can evaluate the specified model weights file on the validation set to verify the model's accuracy. Using PaddleX for model evaluation can be done with a single command:

python main.py -c paddlex/configs/modules/table_recognition/SLANet.yaml \
    -o Global.mode=evaluate \
    -o Global.dataset_dir=./dataset/table_rec_dataset_examples

Similar to model training, the following steps are required:

Specify the .yaml configuration file path for the model (here it's SLANet.yaml)
Specify the mode as model evaluation: -o Global.mode=evaluate
Specify the path to the validation dataset: -o Global.dataset_dir Other related parameters can be set by modifying the Global and Evaluate fields in the .yaml configuration file. For details, refer to PaddleX Common Model Configuration File Parameter Description.

👉 More Details (Click to Expand)

When evaluating the model, you need to specify the model weights file path. Each configuration file has a default weight save path. If you need to change it, simply append the command line parameter to set it, such as -o Evaluate.weight_path=./output/best_accuracy/best_accuracy.pdparams.

After completing the model evaluation, an evaluate_result.json file will be produced, which records the evaluation results, specifically, whether the evaluation task was completed successfully and the model's evaluation metrics, including acc ;

4.4 Model Inference and Model Integration¶

After completing model training and evaluation, you can use the trained model weights for inference predictions or Python integration.

4.4.1 Model Inference¶

Inference predictions can be performed through the command line with just one command. Before running the following code, please download the demo image to your local machine.
```
python main.py -c paddlex/configs/modules/table_recognition/SLANet.yaml  \
    -o Global.mode=predict \
    -o Predict.model_dir="./output/best_accuracy/inference" \
    -o Predict.input="table_recognition.jpg"
```
Similar to model training and evaluation, the following steps are required:
Specify the .yaml configuration file path for the model (here it's SLANet.yaml)
Specify the mode as model inference prediction: -o Global.mode=predict
Specify the model weights path: -o Predict.model_dir="./output/best_accuracy/inference"
Specify the input data path: -o Predict.input="...". Other related parameters can be set by modifying the Global and Predict fields in the .yaml configuration file. For details, refer to PaddleX Common Model Configuration File Parameter Description.
Alternatively, you can use the PaddleX wheel package for inference, easily integrating the model into your own projects.

4.4.2 Model Integration¶

The model can be directly integrated into the PaddleX pipeline or directly into your own project.

1.Pipeline Integration

The table structure recognition module can be integrated into PaddleX pipelines such as the General Table Recognition Pipeline and the Document Scene Information Extraction Pipeline v3 (PP-ChatOCRv3-doc). Simply replace the model path to update the table structure recognition module in the relevant pipelines. For pipeline integration, you can deploy your obtained model using high-performance inference and service-oriented deployment.

2.Module Integration

The model weights you produce can be directly integrated into the table structure recognition module. Refer to the Python example code in Quick Integration , and simply replace the model with the path to your trained model.

You can also use the PaddleX high-performance inference plugin to optimize the inference process of your model and further improve efficiency. For detailed procedures, please refer to the PaddleX High-Performance Inference Guide.