PaddlePaddle / PicoDet_layout_1x

PicoDet_layout_1x

Introduction

A high-efficiency layout area localization model trained on a self-built dataset using PicoDet-1x, capable of detecting 5-Class english document area, including Text, Title, Table, Figure, and List. The key metrics are as follow:

Quick Start

Installation

1. PaddlePaddle

Please refer to the following commands to install PaddlePaddle using pip:

# for CUDA11.8
python -m pip install paddlepaddle-gpu==3.0.0 -i https://www.paddlepaddle.org.cn/packages/stable/cu118/

# for CUDA12.6
python -m pip install paddlepaddle-gpu==3.0.0 -i https://www.paddlepaddle.org.cn/packages/stable/cu126/

# for CPU
python -m pip install paddlepaddle==3.0.0 -i https://www.paddlepaddle.org.cn/packages/stable/cpu/

For details about PaddlePaddle installation, please refer to the PaddlePaddle official website .

1. PaddleOCR

Install the latest version of the PaddleOCR inference package from PyPI:

python -m pip install paddleocr

Model Usage

You can quickly experience the functionality with a single command:

paddleocr layout_detection \
    --model_name PicoDet_layout_1x \
    -i https://cdn-uploads.huggingface.co/production/uploads/63d7b8ee07cd1aa3c49a2026/-oU2IpNLcA0gTMJ34wDBR.png

You can also integrate the model inference of the layout detection module into your project. Before running the following code, please download the sample image to your local machine.

from paddleocr import LayoutDetection

model = LayoutDetection(model_name="PicoDet_layout_1x")
output = model.predict("-oU2IpNLcA0gTMJ34wDBR.png", batch_size=1, layout_nms=True)
for res in output:
    res.print()
    res.save_to_img(save_path="./output/")
    res.save_to_json(save_path="./output/res.json")

After running, the obtained result is as follows:

{'res': {'input_path': '-oU2IpNLcA0gTMJ34wDBR.png', 'page_index': None, 'boxes': [{'cls_id': 0, 'label': 'Text', 'score': 0.9922339916229248, 'coordinate': [84.18926, 1248.3407, 583.2205, 1582.3639]}, {'cls_id': 0, 'label': 'Text', 'score': 0.991148829460144, 'coordinate': [606.0175, 1209.0277, 1105.669, 1538.559]}, {'cls_id': 0, 'label': 'Text', 'score': 0.989486813545227, 'coordinate': [84.11199, 710.821, 582.4992, 1208.4802]}, {'cls_id': 0, 'label': 'Text', 'score': 0.9823781847953796, 'coordinate': [605.9509, 951.964, 1105.8907, 1089.925]}, {'cls_id': 3, 'label': 'Table', 'score': 0.98123699426651, 'coordinate': [608.91614, 301.1981, 1111.2635, 805.45233]}, {'cls_id': 4, 'label': 'Figure', 'score': 0.971710741519928, 'coordinate': [84.329, 200.35394, 577.5106, 681.766]}, {'cls_id': 0, 'label': 'Text', 'score': 0.950641930103302, 'coordinate': [607.27435, 826.2146, 1107.0017, 911.43726]}, {'cls_id': 0, 'label': 'Text', 'score': 0.9468774795532227, 'coordinate': [605.6758, 197.66481, 1105.3107, 265.3606]}, {'cls_id': 1, 'label': 'Title', 'score': 0.86313396692276, 'coordinate': [606.6188, 1170.0265, 827.1196, 1192.0656]}, {'cls_id': 1, 'label': 'Title', 'score': 0.8427881002426147, 'coordinate': [605.2791, 1131.709, 823.0802, 1152.363]}, {'cls_id': 0, 'label': 'Text', 'score': 0.5990690588951111, 'coordinate': [625.3928, 1558.7576, 849.12946, 1583.5391]}]}}

The visualized image is as follows:

For details about usage command and descriptions of parameters, please refer to the Document .

Pipeline Usage

The ability of a single model is limited. But the pipeline consists of several models can provide more capacity to resolve difficult problems in real-world scenJust a few lines of code can experience the inference of the pipelinearios.

PP-TableMagic (table_recognition_v2)

The General Table Recognition v2 pipeline (PP-TableMagic) is designed to tackle table recognition tasks, identifying tables in images and outputting them in HTML format. PP-TableMagic includes the following 8 modules:

• Table Structure Recognition Module
• Table Classification Module
• Table Cell Detection Module
• Text Detection Module
• Text Recognition Module
• Layout Region Detection Module (optional)
• Document Image Orientation Classification Module (optional)
• Text Image Unwarping Module (optional)

You can quickly experience the PP-TableMagic pipeline with a single command.

paddleocr table_recognition_v2 -i https://cdn-uploads.huggingface.co/production/uploads/63d7b8ee07cd1aa3c49a2026/-oU2IpNLcA0gTMJ34wDBR.png \
    --layout_detection_model_name PicoDet_layout_1x \
    --use_doc_orientation_classify False \
    --use_doc_unwarping False \
    --save_path ./output \
    --device gpu:0 

You can also integrate the PP-TableMagic pipeline into your project. Before running the following code, please download the sample image to your local machine.

from paddleocr import TableRecognitionPipelineV2

pipeline = TableRecognitionPipelineV2(
    layout_detection_model_name=PicoDet_layout_1x,
    use_doc_orientation_classify=False,   # Use use_doc_orientation_classify to enable/disable document orientation classification model
    use_doc_unwarping=False, # Use use_doc_unwarping to enable/disable document unwarping module
    device="gpu:0", # Use device to specify GPU for model inference
    )

output = pipeline.predict("-oU2IpNLcA0gTMJ34wDBR.png")
for res in output:
    res.print() ## Print the predicted structured output
    res.save_to_img("./output/")
    res.save_to_xlsx("./output/")
    res.save_to_html("./output/")
    res.save_to_json("./output/")

The default model used in pipeline is PP-DocLayout-L , so it is needed that specifing to PicoDet_layout_1x by argument layout_detection_model_name . And you can also use the local model file by argument layout_detection_model_dir . For details about usage command and descriptions of parameters, please refer to the Document .

Links

PaddleOCR Repo

PaddleOCR Documentation