TextZoom
TextZoom copied to clipboard
A super-resolution dataset of paired LR-HR scene text images
Hi, Thank you for sharing the code the dataset. Could you please tell me how many LR-HR image pairs in total in TextZoom? Is (17367+1619+1411+1243) as reported in Table 1...

wenjia,你好,我想参照你的方法自己制作数据集,用长焦距拍摄图像做HR,短焦距做LR。请问你是如何预处理图像的,具体是如何crop小焦距和大焦距图像的同一个区域且大小一样来做成图像对的?非常期待您的回复,谢谢
Or is it shared somewhere like a google drive?
Hi, could you please provide with a scirpt of needed packages and some information about training, like the version of PyTorch? I am also curious about the hardware(GPU) of your...
Did you train the TSRN model in your paper with a batch size of 512, right? Thank you for your help!