Daesung Kim

Same issue with =, ~|=|, and # ~|#|. I think this is a recent issue. Edit: I could only reproduce this on local hosting (bundle exec jekyll serve). Online website...

I think you can directly feed the trimmed texts to the recognition model, are there reasons for you to pass the texts to the CRAFT model?

https://github.com/clovaai/deep-text-recognition-benchmark/issues/210#issuecomment-883112527 Can you refer to this issue and see if the problem yet persists?

Which language are you interested? For Korean characters, aihub.or.kr would be a great place for you. I am not sure about English.

@Furyboyy Thanks! This temporarily solves the problem, but I am not sure if it's an appropriate solution..

To my best understanding, the authors have excluded the two slides due to incorrect annotations. Please refer to the supplemental paper at https://openaccess.thecvf.com/content/CVPR2022/supplemental/Zhang_DTFD-MIL_Double-Tier_Feature_CVPR_2022_supplemental.pdf.

You're welcome! I do not have an answer for that; hope you hear from the authors soon.

@jihyeheo While I'm not the author, it is highly unlikely. You probably need your own toy dataset :p

> @Treeboy2762 ㅠ 너무하시네 ㅠ.ㅠ 아무래도 듀크대 병원데이터라 공개하긴 어려울거에요.. 저도 조만간 실험해볼 생각이라 만들면 공유해드릴게요 ㅋㅋㅋㅋ

I want to join, but do I need a major-level background in Chemistry / Materials science?