Additional four corruptions
ildoonet opened this issue · 2 comments
ildoonet commented
https://github.com/hendrycks/robustness/blob/master/ImageNet-C/imagenet_c/imagenet_c/__init__.py#L5
In your papers (robustness, augmix), you used 15 corruptions.
Here, the above line of the code, you can see 19 corruptions listed in a tuple.
This might be un-clear to many people, including me. The last four corruptions should be removed so that one can test a model with corruptions as listed in your paper.
hendrycks commented
Those are validation corruptions.
…On Mon, May 18, 2020 at 7:23 PM Ildoo Kim ***@***.***> wrote:
https://github.com/hendrycks/robustness/blob/master/ImageNet-C/imagenet_c/imagenet_c/__init__.py#L5
In your papers (robustness, augmix), you used 15 corruptions.
Here, the above line of the code, you can see 19 corruptions listed in a
tuple.
This can be confusing to many people, including me. Last four corruptions
should be removed so that one can test a model with corruptions as listed
in your paper.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#36>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACZBITULUJNT3QQEOIQAUQLRSHUT3ANCNFSM4NES3MDQ>
.
ildoonet commented
@hendrycks Thanks for the quick reply. After listening to you, I checked the paper again and understood it. But still, it is might be confusing anyone. I think you need to add a comment or split the list into multiples.
Now that I know why there are 19 corruptions, I'm closing this issue. Thanks!