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

Training on custom data #32

Open
SURABHI-GUPTA opened this issue Jan 22, 2022 · 0 comments
Open

Training on custom data #32

SURABHI-GUPTA opened this issue Jan 22, 2022 · 0 comments

Comments

@SURABHI-GUPTA
Copy link

@ppliuboy Thanks for your wonderful work.
I have some doubts regarding self-supervised training for flow prediction.

  1. How to train your model on custom datasets, say youtube videos that do not have any GT annotations? What will be the directory structure?
  2. Regarding the occlusions, say if I create a fixed rectangle to create occlusion, will this method work? Like, which kind of occlusions does your paper talk about.
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