search

aleksandrkim61 / EagerMOT

Official code for "EagerMOT: 3D Multi-Object Tracking via Sensor Fusion" [ICRA 2021]
MIT License
404 stars 78 forks source link

Does the code require any training? #5

Closed sisaha9 closed 3 years ago

sisaha9 commented 3 years ago

I am just catching up on this work. I don't see a training script set up anywhere. Can this code be modified to run real time instead of a dataset without any training beforehand?

sisaha9 commented 3 years ago

Ah I see it relies on pre established boundary boxes. So we would provide that in a similar format. Thanks!

aleksandrkim61 commented 3 years ago

You got it! The fact that tracking is not coupled to any specific input source or data is one of its big advantages.

You do not even have to provide detections in the same exact format, you can just add a custom parser to inputs/detections_{3d/2d} to consume your format. That's why there are so many parsing functions - to accommodate different detection formats provided by the community that we used in our experiments.