Synthetic data

[davasli]

Hi,

Thanks for sharing your code. This is a great paper. I was able to run your code without any errors and can see the folders and images created. It trained very quickly. However, the synthetic data looks exactly the same as the real data. I'm running the "coco10_cat_pipe.sh" example for coco with 10 samples. I have the following files:

• mix_n333_HED_p512_prcat_dfsNone_seed1_imprior
• mix_n333_HED_p512_prcat_dfsNone_seed1_imprior_avgacsl30
• syn_n350_HED_p512_prcat_dfsNone_seed1_imprior

Thanks for your help in advance,
Kemal

[ghostgeek6113]

Hi,

Thanks for sharing your code. This is a great paper. I was able to run your code without any errors and can see the folders and images created. It trained very quickly. However, the synthetic data looks exactly the same as the real data. I'm running the "coco10_cat_pipe.sh" example for coco with 10 samples. I have the following files:

• mix_n333_HED_p512_prcat_dfsNone_seed1_imprior
• mix_n333_HED_p512_prcat_dfsNone_seed1_imprior_avgacsl30
• syn_n350_HED_p512_prcat_dfsNone_seed1_imprior

Thanks for your help in advance, Kemal

Hello Davasli,

I hope you are well. I have been trying to run the code but I am facing some configuration challenges especially in the file structure. I was wondering if you had some time to spare and share with me your process or any unusual thing you discovered.
Also when you said trained very quickly, how fast?

Hoping to hear from you soon,
Thank you