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

Weak result when using my own video #23

Open
ahmadi3d opened this issue Mar 7, 2021 · 1 comment
Open

Weak result when using my own video #23

ahmadi3d opened this issue Mar 7, 2021 · 1 comment

Comments

@ahmadi3d
Copy link

ahmadi3d commented Mar 7, 2021

Hello
I just downloaded a traffic video from YouTube :
https://www.youtube.com/watch?v=mrjqS5zd268&ab_channel=TaverNet
I gave these values as ground truth :
258.0000,698.0000,388.0000,201.0000
When I run the tracker it starts great but as soon as a very small occultation happens it will lose the target and keeps changing targets! if you try the sample you can see what I mean
Reading the paper I understood the process and it shouldn't be so weak against occultation and disappearances, I don't understand why something like this is happening here, maybe its a bug ?, I would really appreciate it if you could check it out and tell me what I can do to fix this, thank you.

@pvoigtlaender
Copy link
Contributor

I think this sequence is just extremely hard due to the high object density.
You can try to tweak the hyper-parameters, especially the weighting factors of the TDPA.

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