Table Detection And Extraction

🚀 YOLOv8s Table Detection Model

This model is based on the YOLO framework, designed to detect tables in images, whether they are bordered or borderless. It has high accuracy in table detection and can distinguish between different types of tables.

🚀 Quick Start

Installation

To start using the YOLOv8s Table Detection model, install the necessary libraries with the following command:

pip install ultralyticsplus==0.0.28 ultralytics==8.0.43

Load Model and Perform Prediction

from ultralyticsplus import YOLO, render_result

# load model
model = YOLO('foduucom/table-detection-and-extraction')

# set model parameters
model.overrides['conf'] = 0.25  # NMS confidence threshold
model.overrides['iou'] = 0.45  # NMS IoU threshold
model.overrides['agnostic_nms'] = False  # NMS class-agnostic
model.overrides['max_det'] = 1000  # maximum number of detections per image

# set image
image = '/path/to/your/document/images'

# perform inference
results = model.predict(image)

# observe results
print(results[0].boxes)
render = render_result(model=model, image=image, result=results[0])
render.show()

✨ Features

• Precise Table Detection: The model can accurately detect both bordered and borderless tables in images.
• Data Extraction: Integrated with OCR technology, it can extract data from detected tables.
• Versatile Application: Can be fine - tuned for specific tasks or integrated into larger applications.

📦 Installation

pip install ultralyticsplus==0.0.28 ultralytics==8.0.43

💻 Usage Examples

Basic Usage

from ultralyticsplus import YOLO, render_result

# load model
model = YOLO('foduucom/table-detection-and-extraction')

# set image
image = '/path/to/your/document/images'

# perform inference
results = model.predict(image)

# observe results
print(results[0].boxes)
render = render_result(model=model, image=image, result=results[0])
render.show()

Advanced Usage

from ultralyticsplus import YOLO, render_result

# load model
model = YOLO('foduucom/table-detection-and-extraction')

# set model parameters
model.overrides['conf'] = 0.25  # NMS confidence threshold
model.overrides['iou'] = 0.45  # NMS IoU threshold
model.overrides['agnostic_nms'] = False  # NMS class-agnostic
model.overrides['max_det'] = 1000  # maximum number of detections per image

# set image
image = '/path/to/your/document/images'

# perform inference
results = model.predict(image)

# observe results
print(results[0].boxes)
render = render_result(model=model, image=image, result=results[0])
render.show()

📚 Documentation

Model Summary

The YOLOv8s Table Detection model is an object - detection model based on the YOLO framework. It can detect tables in images and distinguish between bordered and borderless ones.

Model Details

Model Description

The model can precisely identify tables in images, whether bordered or borderless. It is also useful for unstructured document analysis. By using bounding boxes, it helps users isolate tables. Integrated with OCR, it can extract data from tables, streamlining information retrieval from unstructured documents.

• Developed by: FODUU AI
• Model type: Object Detection
• Task: Table Detection (Bordered and Borderless)

Supported Labels

['bordered', 'borderless']

Uses

Direct Use

The model can directly detect tables in images and distinguish between bordered and borderless tables.

Downstream Use

It can be fine - tuned for specific table - detection tasks or integrated into larger applications for furniture recognition, interior design, image - based data extraction, etc.

Out - of - Scope Use

The model is not for unrelated object - detection tasks or scenarios outside table detection.

Bias, Risks, and Limitations

• Performance may vary based on training data quality, diversity, and representativeness.
• It may face challenges in detecting tables with intricate designs or complex arrangements.
• Accuracy may be affected by lighting conditions, image quality, and resolution.
• Detection of very small or distant tables may be less accurate.
• Classification of bordered and borderless tables may be influenced by design variations.

Recommendations

Users should be aware of the model's limitations and potential biases. Further testing and validation are recommended for specific use cases.

Training Details

Training Data

The model is trained on a diverse dataset with table images from various sources, including both bordered and borderless tables.

Training Procedure

The training involves extensive computation over multiple epochs. Model weights are adjusted to minimize detection loss and optimize performance.

Metrics

Model Architecture and Objective

The YOLOv8s architecture uses a modified CSPDarknet53 as its backbone, along with self - attention mechanisms and feature pyramid networks, enabling accurate table detection and classification.

Compute Infrastructure

Hardware

NVIDIA GeForce RTX 3060 card

Software

The model was trained and fine - tuned using a Jupyter Notebook environment.

🔧 Technical Details

The YOLOv8s architecture employs a modified CSPDarknet53 as its backbone. Self - attention mechanisms and feature pyramid networks are used to enhance the model's ability to detect and classify tables accurately, considering variations in size, design, and style.

📄 License

No license information provided in the original document.

Model Card Contact

For inquiries and contributions, please contact us at info@foduu.com.

@ModelCard{
  author    = {Nehul Agrawal and
               Pranjal Singh Thakur},
  title     = {YOLOv8s Table Detection},
  year      = {2023}
}