DocGeoNet

DocGeoNet copied to clipboard

About dataset license

3

Hi, The DocGeoNet code repo is under MIT license, but I don't see a license for the DIR300 dataset. Can I use the DIR300 dataset to train a model for...

Soongja

Hi! The feature visualization in Figure 6 is cool.  Could you provide some details about how to reproduce them?

who3ohw

Questions about data processing while training

1

Hi, I noticed that you use the absolute value of bm while training and do not use normalization. ``` bm = bm_ori / 448 * 255. ``` Have you tried...

KingRicardo

Could you please provide the code to train custom dataset?

WongVi

Questions about data normalization

1

1. For training, is wc(3D coordinate map) normalized in the same way as wc in DewarpNet?([reference](https://github.com/cvlab-stonybrook/DewarpNet/blob/22960ffb9a359ec6b9ccd8d72404bfedd36781be/loaders/doc3dwc_loader.py#L82-L91)) 2. After network inference, why is bm being multiplied by 0.99 as below? https://github.com/fh2019ustc/DocGeoNet/blob/5a56ab427af6748f96908ce089c5ceea0b6dfeed/inference.py#L29

Soongja

Some questions about the textlines annotation process

5

Hello hao, Thanks for your awesome work for document image dewarping. Could you provide more details about the textlines annotation process? (e.g., the kernel size of binarization and dilation, and...

KingRicardo

Questions on OCR evaluation

2

Hi, I have a few questions on OCR evaluation. 1. When evaluating OCR performance on DIR300 dataset(or DocUNet benchmark), the size of the predicted image and GT image are different....

Soongja

OCR指标存在一些diff

3

作者您好： 我们所使用的矫正图均为github上贴的，但是我们评测的OCR指标与paper中的指标存在一些diff，我们的环境是mac os。 这是我们的tesseract以及pytesseract版本： 这是我们复线的指标： 非常感谢您的工作，希望能得到您的解答！

dbylynn

Where can I find paper

2

Thank you for open source your research code. Where can I find the paper? Thanks

Sanster