Method

Online Domain Adaptation for Multi-Object Tracking [on] [ODAMOT]


Submitted on 1 May. 2015 21:29 by
Eleonora Vig (Xerox Research Centre Europe)

Running time:1 s
Environment:1 core @ 2.5 Ghz (Python)

Method Description:
Online Domain Adaptation for Multi-Object Tracking
Parameters:
Latex Bibtex:
@inproceedings{Gaidon2015BMVC,
author = {Gaidon, A. and Vig, E.},
booktitle={British Machine Vision Conference (BMVC)},
title = {{Online Domain Adaptation for Multi-Object Tracking}},
year = {2015}
}

Detailed Results

From all 29 test sequences, our benchmark computes the commonly used tracking metrics CLEARMOT, MT/PT/ML, identity switches, and fragmentations [1,2]. The tables below show all of these metrics.


Benchmark MOTA MOTP MODA MODP
CAR 59.23 % 75.45 % 60.36 % 82.75 %

Benchmark recall precision F1 TP FP FN FAR #objects #trajectories
CAR 65.97 % 94.65 % 77.75 % 23821 1346 12288 12.10 % 26566 2004

Benchmark MT PT ML IDS FRAG
CAR 27.08 % 57.38 % 15.54 % 389 1274

This table as LaTeX


[1] K. Bernardin, R. Stiefelhagen: Evaluating Multiple Object Tracking Performance: The CLEAR MOT Metrics. JIVP 2008.
[2] Y. Li, C. Huang, R. Nevatia: Learning to associate: HybridBoosted multi-target tracker for crowded scene. CVPR 2009.


eXTReMe Tracker