Wrong counts for groundtruth objects when there are no predicted objects in a frame

cheind / py-motmetrics

:bar_chart: Benchmark multiple object trackers (MOT) in Python

MIT License

1.37k stars 259 forks source link

Closed jvlmdr closed 3 years ago

jvlmdr commented 3 years ago

This bug only affects the develop branch. I have added a test to highlight the problem in 02ba3da06a40e3841e54ee1c93d4dc345a69cca3

jvlmdr commented 3 years ago

Fixed in 6597e8a4ed398b9f14880fa76de26bc43d230836