search

VisualComputingInstitute / mots_tools

Tools for evaluating and visualizing results for the Multi Object Tracking and Segmentation (MOTS) task
MIT License
222 stars 45 forks source link

Sometimes it reports error "Something went wrong! FN is negative". Why does it happen? #11

Closed zHaOshuAnGye closed 3 years ago

zHaOshuAnGye commented 3 years ago

Hello~Thanks for your code.It's very nice. When using it , I encountered the above problems. I saw the source code,“tmpfn += len(g) - num_associations”，this line，but I don't know why sometimes len(g) is less than num_associations. I want to ask that under what circumstances can it happen. It shouldn't have happened in theory.

zHaOshuAnGye commented 3 years ago

Maybe it is because gg may match more than one mask in tt? Can I add a "break" at end of this loop, which means gg only match the first mask meeting in tt. image

ahnonay commented 3 years ago

There should be only one predicted mask having IoU above 0.5 with a ground truth mask if all predicted masks are non-overlapping. How are you calling compute_MOTS_metrics_per_sequence? Are you calling it via eval.py?

ahnonay commented 3 years ago

Closing this for now, please reopen if needed