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

do you train your online tracker without public detection? #3

Closed
chenrxi opened this issue Jun 11, 2021 · 1 comment
Closed

do you train your online tracker without public detection? #3

chenrxi opened this issue Jun 11, 2021 · 1 comment

Comments

@chenrxi
Copy link

chenrxi commented Jun 11, 2021

Have you ever trained your online tracker without public detection? If so, how to detect new track in private detector mode and distinguish from the track already existed successfully. Also, looking forward to your code and models available and thanks a lot for your great job.

@songguocode
Copy link
Owner

It is trained in a way similar to training a Faster-RCNN detector, while the rpn is discarded during tracking and RoIs are provided by public detections. It does not actively detect new targets, but refines positions of existing tracked targets given earlier information. A new target is only initiated with a public detection that is not covered by existing targets. Hope this clarifies your questions.

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