search

cheind / py-motmetrics

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

difference between eval_motchallenge and evaluateTracking #123

Open songsonha opened 3 years ago

songsonha commented 3 years ago

Sorry because I have a dummy question. But I don't really understand what is the difference between eval_motchallenge and evaluateTracking module. Can someone answer that for me pls?

  1. question: what is the difference between MOT16/17 and Motchallenge? There are some different purposes for them?
cheind commented 3 years ago

I think I can answer question 1

Good question, actually! Wondered about it myself :) eval_motchallenge was written by myself as a general purpose evaluation for motchallenge metrics. Later evaluateTracking was added, because some MOT challenges (MOT16/17) required special preprocessing of the provided files (removing mostly hidden ground-truth etc.)

See this commit: https://github.com/cheind/py-motmetrics/commit/3f8f5045d3a5b8b83ad907564c8b441edcbb6f7d

Seems like a lot of code got duplicated in that commit. So if you feel like codensing the apps into single application feel free to submit a PR.