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.

Sign up for GitHub

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

Jump to bottom

I have some questions about confidence #9

Open
lzqcode opened this issue Dec 16, 2018 · 2 comments
Open

I have some questions about confidence #9

lzqcode opened this issue Dec 16, 2018 · 2 comments

Comments

@lzqcode
Copy link

lzqcode commented Dec 16, 2018

Why do we judge confidence from the second mat of prob and use "if (confidence_data[i] <= 1 - thresh)"? This means that the smaller the value, the greater the probability of face in this area. But when NMS is used, after sort, the larger bbox of confidence is left in this part. Isn't it contradictory here? I printed the second mat of prob to see that there is a face in the original map. The place is black (small value)
@lzqcode
Copy link
Author

lzqcode commented Dec 16, 2018

Can you leave an email address for better communication? My email address is [email protected]. Thank you.

@szad670401
Copy link
Owner

the softmax output order of pnet model that i trained is different from origin one.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@szad670401 @lzqcode