SciTSR
SciTSR copied to clipboard
Table structure recognition dataset of the paper: Complicated Table Structure Recognition
1. `eval_relations(gt=[ground_truth_relations], res=[your_relations], cmp_blank=True)` - As the objective is to find the merged relations of the neighboring cells, isn't the blank relation comparison a wrong evaluation metric? 2. `rel_gen.py` [this](https://github.com/Academic-Hammer/SciTSR/blob/79954b5143295162ceaf7e9d9af918a29fe12f55/scitsr/data/rel_gen.py#L35)...
Thanks for providing the data. To run your evaluation code, it needs content to be in the structure json files. 1) Do you have a code that measure the structure...
In chunks, the bboxes are only given for filled cells and we can get same during inference. If there are empty cells, how can we get coordinates for those or...
can you provide me script to create file ".rel" ?? Thanks
Some of the images in the dataset are beyond the right edge, but the cells beyond the image edge are still present in the annotation file. There are about 200...
I recovered the structure of the table using chunk(.chunk) and structure(.json) files. When I visualized the annotation, I found that some annotations are wrong.. (not included in scitsr-comp.list) The box...
HI, Thanks for the excellent work. I might be wrong but the coordinates stored in the chunk files don't seem to match the provided image. Could you let me know...
Hi, thanks for providing your code! I am facing problems with chunk ordering while using a different method to generate chunk files. I was wondering if you could explain how...