cheind / py-motmetrics

:bar_chart: Benchmark multiple object trackers (MOT) in Python
MIT License
1.37k stars 259 forks source link

Default value of max_iou in the function iou_matrix #137

Open Sergigb opened 3 years ago

Sergigb commented 3 years ago

Hi!

I have a little doubt about the iou_matrix function. According its description, max_iou is set to 0.5 by default (which is the value mentioned in the MOT16 challenge article if I'm not mistaken). However, the definition of the function actually defaults the value to 1:

def iou_matrix(objs, hyps, max_iou=1.):

Shouldn't the default value of max_iou be 0.5, since this is the described behaviour in the MOT16 challenge? Having it set to 1 makes any tiny overlap count as a match, which doesn't make much sense to me.

Thanks!