search

jessemelpolio / non-stationary_texture_syn

Code used for texture synthesis using GAN
http://vcc.szu.edu.cn/research/2018/TexSyn
MIT License
369 stars 66 forks source link

About experiments in your paper #8

Closed wzm2256 closed 5 years ago

wzm2256 commented 6 years ago

Hi, I have a question that is not directly related to your algorithm. It's about the experiments in your paper.

I notice you compared your result with Self-tuning and MGANs in figure 7. As far as I know, the code for MGANs can not be used for texture synthesis, although they have these results in their paper, and there are some bugs in the code of self-tuning methods. How do you run their codes?

jessemelpolio commented 5 years ago

I just download their codes on github and make several minor changes without changing the network architecture and main ideas. Seeing that our results show similar characteristics with the results posted in their paper, we think those results are correct to some extent. Once in a while, we have the similar doubts with you (also the reviewers have pointed out this problem) and we carefully checked the code and just in case, we turned to another student to perform the same experiments based on their code. But similar results are achieved. Sorry that I forgot all the details for such a long time.

wzm2256 commented 5 years ago

Thank all the same.