mega.pytorch icon indicating copy to clipboard operation
mega.pytorch copied to clipboard

Published 20 hours ago • verified publisher icon Scalsol

How to train on ICDAR2013 dataset?

Open cyborg1995 opened this issue 3 years ago • 0 comments

Hello, I want to train on the ICDAR2013 dataset.

The dataset consists of video files as opposed to image sequences in ImageNet. How do I convert it to the required format? Also, the annotations are in .XML files.

CUSTOMIZE.md mentions that I should prepare a txt file, how do I prepare it from the dataset that I have?

cyborg1995 avatar Jun 20 '21 07:06 cyborg1995