jiananli2016 / layoutgan-tensorflow Goto Github PK

Usage
First, download the trasformed point layout representation of MNIST dataset from
https://drive.google.com/file/d/1R1iRZxADR_RcDsuR4gyStyLAo7i5LRAH/view?usp=sharing,
and put it under ./data directory.

hi, sir, the url of the google cloud can't be accessed, could you pleased to check it, thank you anyway.

could you fix this link failure? or just share the files for data preprocessing?

Hello,
Can you please share the link for tangram dataset?

Thanks

Could you share your scripts for generating pre_data_cls.npy file or share the pre_data_cls.npy file since the google driver link is failed?

Hello, could you please share your checkpoints? Thanks a lot.

First, download the trasformed point layout representation of MNIST dataset from
https://drive.google.com/file/d/1x_cXGILI5EQvN7ksUH7Kz8-VCPsGnq_S/view?usp=sharing,
and put it under ./data directory.

Got the following error:

Sorry, the file you have requested does not exist.

Make sure that you have the correct URL and the file exists.

Hi, Jianan. Thanks for sharing the code.
I am trying to reproduce the bbox layout experiment in my forked repository.
Referring to the paper and the commented code, I obtained the following generated results.

Real example	Generated example

Do you think the quality of these generated results is as good as the results you got?
Do you expect improvement by adjusting hyperparameters? Which parameters in particular do you feel have a significant impact?
Also, I would like to know if there are any significant differences in my implementation of the NN architecture.

My experiment is summarized here.