-
http://people.idsia.ch/~juergen/icdar_2007.pdf
-
Hi, dengdan, thank you for your hard work,
i am trying to train seglink model on my own datasets, i meet such situation:
1. I only get one GPU card which is TITAN XP, i have rewrite your train scrip…
-
Hi,
Thanks for the code.
I've tried to reproduce the reported results for ICDAR 2013 and ICDAR 2015.
Using the weights you've shared, I get .7 F1 on ICDAR 2013 E2E. In the paper you reported…
-
But here it is being written it is just the detector , what about recognition. I havent run it for now but i wanna know whether this code is capable of recognition or not.
If not then came someone …
-
CTPN's performance on ICDAR 2013 dataset is F-measure 0.88 (R 0.83, P 0.93), not 0.822 in your paper (table 2 on page 11.) They published test model at https://github.com/tianzhi0549/CTPN (downloadabl…
jwnsu updated
6 years ago
-
so nice to share the code here.
I have a question, the text bounding box may be incline in one image. so to determine a inline bounding box, (xmin, ymin, xmax, ymax) is not enough, for example, we ma…
-
Thanks for sharing the paper and repository. Wondering where to download your pretrained model file so that we can try with ICDAR 13 and 15 dataset. A pointer to training dataset will be great too.
jwnsu updated
6 years ago
-
论文中提到，将ICDAR2015将短边resize到900，ICDAR2013将短边resize到640，这是靠反复实验得到的一个参数吗？或者说有没有什么选择的方法？作者有没有考虑过multi-scale的测试？我换了数据，loss在训练一千次之后，一直到十万次，数值一直是在零点几到二点几之间波动，learning rate已经设置到1*e-5，感觉一直没有收敛，请问作者在之前实验过程中遇到过吗？
-
[ LAREX - A semi-automatic open-source Tool for Layout Analysis and Region Extraction on Early Printed Books](https://arxiv.org/abs/1701.07396)
[论文下载](https://arxiv.org/ftp/arxiv/papers/1701/1701…
-