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 meaning of hthresh and othresh in the Generate_detection.py #59

Closed
whqwill opened this issue Nov 6, 2020 · 5 comments
Closed

Comments

@whqwill
Copy link

whqwill commented Nov 6, 2020

what's the meaning of hthresh and othresh in the Generate_detection.py in instance level

@whqwill
Copy link
Author

whqwill commented Nov 6, 2020

BTW, where is the generation_args.pkl from?

@Foruck
Copy link
Collaborator

Foruck commented Nov 8, 2020

  1. It means the human and object instances with detection confidence lower hthresh and othresh are ignored in inference and evaluation.
  2. They are obtained via grid search on selected validation set from HAKE. You could take reference from Wired results: obtain the similar reported result without training... #30 (comment).

@whqwill
Copy link
Author

whqwill commented Nov 8, 2020

  1. It means the human and object instances with detection confidence lower hthresh and othresh are ignored in inference and evaluation.
  2. They are obtained via grid search on selected validation set from HAKE. You could take reference from #30 (comment).

OK. Thanks. So the hthresh and othresh are also obtained via grid search on selected validation set from HAKE, right?

@whqwill
Copy link
Author

whqwill commented Nov 8, 2020

BTW, if I want to do inference on my own dataset, how should I set hthresh and othresh and other parameters?

@Foruck
Copy link
Collaborator

Foruck commented Nov 12, 2020

For Q1, that's right.
For Q2, maybe a grid search on a selected validation set from your own dataset could do it.

@whqwill whqwill closed this as completed Nov 13, 2020
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

2 participants