textbsr

textbsr copied to clipboard

no ckpt file

6

no ckpt file is found in /home/cooper/.cnstd/1.2/db_resnet34 It seems that some files are missing, how can I download them?

dikapiliao1

Hi, thanks for releasing this nice project. Some of the code ([L58](https://github.com/csxmli2016/textbsr/blob/1e5b82092ae2bbc363d7e26a2f9f01cc19a9010d/textbsr/models/TextEnhancement.py#L58C7-L58C7) ~ [L85](https://github.com/csxmli2016/textbsr/blob/1e5b82092ae2bbc363d7e26a2f9f01cc19a9010d/textbsr/models/TextEnhancement.py#L85) from TextEnhancement.py) was hard for me to understand. Could you explain what `move_w, move_h, ref_point` are...

649459021

Does textbsr support Japanese or other languages?

7

Does textbsr support Japanese or other languages?

Ken1256

Can provide x1 and x2 pre-trained model?

3

Can provide x1 and x2 pre-trained model? Thanks

Ken1256

关于整图文本超分

6

作者您好，先前的工作都是基于单行文本图像进行超分的，我想知道现在是如何做到整图文本图像进行超分的呢？思路是怎样的？感谢回复

renjd

no ckpt file is found in /home/work/.cnstd/1.2/db_resnet34

1

EchoXu98

train dataset

3

Can the dataset of this model be published? If not, can you tell me how to build the dataset, including the MARCONet you mentioned in the Readme?

HW140701

improving the bsrgan model

3

is it possible retrain the model with [dat](https://github.com/zhengchen1999/DAT) architecture ? It could give better results than bsrgan

fgogrrore4ft3w

How can I upscale the entire image with the model without cnstd

1

cnstd doesn't detect the small text , is it possible to bypass text detection and just upscale the entire image with textbsr?

junmittens2004