Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

What's the resolution of videos when eval on GoPro dataset in your paper? And any advice about the resolution when doing video debluring? #48

Open
double-fire-0 opened this issue Sep 1, 2022 · 0 comments

Comments

@double-fire-0
Copy link

In Table 2 of your paper, you report the result of your model on GoPro dataset. At what resolution did you get this result?

And can you give some advice about the resolution when doing video debluring?

From this issue #12 , it seems that this mode is hard to run at a high resolution.
And 180p is blurry in application, how can we apply this model on a higher resolution (eg.720p)
Is it possible for us to train this model on a lower resolution and inference on a higher resulution?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant