Siamese Multi Object Tracker


Short name:

SiamMOT

Benchmark:

Description:

Siamese Network is used for visual cue while IOU net is used as location and scale cues. Total score is the weighted sum of the scores of the cues. Location prediction is implemented to estimate the location of each target in next frame based on the trajectory. Single Object Tracking scheme is employed to recover missing detection.

Hardware:

i7-7700K 4.2GHz, 4 Core, Titan xp

Detector:

Public

Processing:

Online

Last submitted:

June 13, 2019 (5 months ago)

Published:

June 16, 2019 at 00:00:00 CET

Submissions:

3

Open source:

No

Project page / code:

n/a

Reference:

Anonymous submission

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
37.677.514.517.7 % 23.0 % 64,916275,0729,70911,831i7-7700K 4.2GHz, 4 Core, Titan xpPublic
IDF1ID PrecisionID Recall
27.635.922.4

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
CVPR19-0451.631.680.96.269221.7 % 16.6 % 12,936135,3415,7076,311
CVPR19-0620.021.270.926.72639.5 % 30.0 % 26,87875,7102,3033,162
CVPR19-0748.939.675.84.811134.2 % 14.4 % 2,81713,602488610
CVPR19-084.517.169.227.61904.7 % 41.6 % 22,28550,4191,2111,748

Raw data:

n/a


SiamMOT