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

Hard Negative Mining problem #121

Open
shinji-ohkubo opened this issue Nov 18, 2020 · 0 comments
Open

Hard Negative Mining problem #121

shinji-ohkubo opened this issue Nov 18, 2020 · 0 comments

Comments

@shinji-ohkubo
Copy link

multibox_loss.py
-->num_neg = torch.clamp(self.negpos_ratio * num_pos, max=pos.size(1) - 1)

If positive is 0, negative is also 0, which hinders negative learning and causes over-detection. I think it is correct to set a fixed minimum value.

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