Joint Detection, Identification and Tracking of Multiple Objects in an Online Manner

MOT16-01 MOT16-03 MOT16-06 MOT16-07 MOT16-08 MOT16-12 MOT16-14

Short name:

Joint_DIT

Benchmark:

Description:

In this work, we present a unified approach that simultane-ously detects, identifies and tracks objects in two consecutive frames ofa given video for the task of Multiple Object Tracking.

Hardware:

Intel Xeon CPU 2.6GHz 256GB RAM

Detector:

Private

Processing:

Online

Last submitted:

July 25, 2018 (7 months ago)

Published:

July 07, 2018 at 02:00:24 CET

Submissions:

2

Open source:

No

Project page / code:

n/a

Reference:

Anonymous submission

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
43.078.31.214.0 % 43.5 % 7,02596,1846372,866Intel Xeon CPU 2.6GHz 256GB RAMPrivate
IDF1ID PrecisionID Recall
42.162.231.8

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
MOT16-0137.037.274.50.22317.4 % 47.8 % 1123,89127114
MOT16-0353.847.178.72.114819.6 % 20.9 % 3,16944,7523771,609
MOT16-0653.642.076.80.222119.0 % 34.4 % 2885,00064268
MOT16-0724.635.177.10.8547.4 % 63.0 % 40411,85349314
MOT16-0828.033.280.60.66312.7 % 46.0 % 40211,60241135
MOT16-1229.843.779.71.08611.6 % 55.8 % 9334,8751169
MOT16-1413.522.674.02.31645.5 % 61.6 % 1,71714,21168357

Raw data:

n/a


Joint_DIT
TDP