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

How to keep the bounding boxes when interval is NOT zero? #604

Open
lida2003 opened this issue Dec 14, 2024 · 3 comments
Open

How to keep the bounding boxes when interval is NOT zero? #604

lida2003 opened this issue Dec 14, 2024 · 3 comments

Comments

@lida2003
Copy link

Interval can speed up FPS data, but the bounding box is blinking also.

How to keep the box in every frame? Kind of tracking ability?

@SarperYurttas
Copy link

Have you added the nvtracker plugin to your pipeline? The NvDCF tracker is good at dealing with inference intervals of more than 0.

https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_plugin_gst-nvtracker.html#low-level-tracker-comparisons-and-tradeoffs

@lida2003
Copy link
Author

lida2003 commented Jan 6, 2025

@SarperYurttas I think I didn't try this yet.
So do you imply that the NvDCF tracker has implemented drawing those bbox if inference is NOT used?
I asked on nvidia forum also about this issue, no luck yet.
But I'll try your idea to see if NvDCF has implemented the functionality to draw bounding boxes for frames other than every nth inference frame.

@SarperYurttas
Copy link

In the documentation, here it says:

NvDCF: The NvDCF tracker is an online multi-object tracker that employs a discriminative correlation filter for visual object tracking, which allows independent object tracking even when detection results are not available. It uses the combination of the correlation filter responses and bounding box proximity for data association.

So if your inference engine misses the detection(False negative) or it has an interval, the tracker looks for possible locations for tracked objects and generates a bounding box if its confidence is higher than the threshold that you specified in the config file. It's of course way more complicated than that so please read the documentation carefully to understand what is going on exactly.

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