/TableNet

Unofficial implementation of "TableNet: Deep Learning model for end-to-end Table detection and Tabular data extraction from Scanned Document Images"

Primary LanguageJupyter Notebook

TableNet

Unofficial implementation of ICDAR 2019 paper : TableNet: Deep Learning model for end-to-end Table detection and Tabular data extraction from Scanned Document Images.

Paper

Overview

Paper: TableNet: Deep Learning model for end-to-end Table detection and Tabular data extraction from Scanned Document Images

TableNet is a modern deep learning architecture that was proposed by a team from TCS Research year in the year 2019. The main motivation was to extract information from scanned tables through mobile phones or cameras.

They proposed a solution that includes accurate detection of the tabular region within an image and subsequently detecting and extracting information from the rows and columns of the detected table.

Architecture: The architecture is based out of Long et al., an encoder-decoder model for semantic segmentation. The same encoder/decoder network is used as the FCN architecture for table extraction. The images are preprocessed and modified using the Tesseract OCR.

Source: Nanonets

architecture

How to run

pip install -r requirements.txt
  1. Download the Marmot Dataset from the link given in readme.
  2. Run data_preprocess/generate_mask.py to generate Table and Column Mask of corresponding images.
  3. Follow the TableNet.ipynb notebook to train and test the model.

Challenges

  • Require a very decent System with a good GPU for accurate result on High pixel images.

Dataset

Download the dataset provided in paper : Marmot Dataset.