search

qc17-THU / DL-SR

Tensorflow/keras implementation for image transformation from low-resolution (LR) image to super-resolved one, including single wide-field (WF) image super-resolution prediction and SIM reconstruction.
MIT License
98 stars 28 forks source link

python version of your matlab code. process *.mrc #9

Closed sbelharbi closed 2 years ago

sbelharbi commented 2 years ago

hi, could you please release the python code that does the job of matalb code you provided here? or release the final tif images produced by your matlab code? for full data. the python code you provided with data in supp material does read/write only. thanks

qc17-THU commented 2 years ago

Hi, we didn't have python codes for data augmentation, we performed all the preprocessing, post-processing and evaluation on MATLAB. But in fact if you could read .mrc file via the provided python code in our Figshare page, the following part of data augmentation should be quite simple to write.

sbelharbi commented 2 years ago

hi, thanks. is it possible to provide directly the augmented data? it would be much more helpful. it avoids the need for matlab, or re-writing code. also, it allows other users to train and most importantly evaluate on the same data. thanks

can you give a look into the matlab issue? it runs fine on one folder f-actin. but it gives errors on other folders. we appreciate your help and sharing the data. thanks

qc17-THU commented 2 years ago

The augmented training dataset contains over 40,000 files, which is extremely slow to upload from my workstation. We have tested that using the augmented data by the MATLAB codes could reproduce the results well in our paper.

I have answered the other problem in issue #10.

sbelharbi commented 2 years ago

thanks. we are trying to run the new matlab code. but it seems to have another issue over f-actin non-linear. by the way, you can compress train and valid sets (40000 + testset files) into a single large file, and dump it somewhere. it may take some time to upload, but once done, the issues with data could be resolved. you can give it the necessary time to upload. we really appreciate your help.

thanks again

qc17-THU commented 2 years ago

See issue #13