Which dataset is used for training

DevashishPrasad / CascadeTabNet

This repository contains the code and implementation details of the CascadeTabNet paper "CascadeTabNet: An approach for end to end table detection and structure recognition from image-based documents"

MIT License

1.47k stars 422 forks source link

Which dataset is used for training #59

Closed shunnnosukekawai closed 3 years ago

shunnnosukekawai commented 3 years ago

I am trying to retrain for ICDAR 19 (Track B2 Modern) table structure recognition, with the latest pytorch 1.5.1 and mmdetection 2.2.1, so I would like to know which dataset is used in this dataset link (https://drive.google.com/drive/folders/1RuEACireEqPnQrYqghEqsOPRE-CkGSUd).

And I would be happy if you could teach me how to fix the config file (cascade_mask_rcnn_hrnetv2p_w32_20e.py) for mmdetection 2.2.1.

DevashishPrasad commented 3 years ago

Iterative Transfer Learning (ITL) was used to train "ICDAR 19 (Track B2 Modern) table structure recognition" model. Details about ITL are mentioned in the paper.

First the model was trained on General dataset and then fine-tuned on ICDAR Table cells dataset

kkissmart commented 3 years ago

where did you use the tablebank?