Obtaining and preparing the dataset. about tfd-icdar2019

[MaliParag]

Hi,

1. You need to download articles on your own as we can not provide them because of copyright issues.

2. Please download the articles using text files in the EasyDownload directory

"GTDB_18files.txt": These 18 documents can be downloaded easily from command line : wget -P [outputDir] -i GTDB_18files.txt

"GTDB_29files.txt": for the rest of the articles you need to check the URLs and download them from there.

I will probably add the name the file should be renamed to. But, for now, use common sense as there is no clash in the author's name and journal year.

3. TFD-ICDAR datasets in this repository provide bounding boxes for the math expressions. I do not know of any other dataset that provides bounding boxes for both embedded and displayed math expressions. I will update here if I find one.

[jjdredd]

Thanks much for the reply. I think I have dealt with the issue 2 successfully with scripts:

https://github.com/jjdredd/TFD-ICDAR2019/blob/master/TFD-ICDAR2019v2/file_download_names_18.script.txt

https://github.com/jjdredd/TFD-ICDAR2019/blob/master/TFD-ICDAR2019v2/file_download_names_29.script.txt