Spatial-Temporal Relation Networks for Multi-Object Tracking


Short name:

STRN_MOT16

Benchmark:

Description:

Spatial-Temporal Relation Networks for Multi-Object Tracking, ICCV2019,
https://arxiv.org/abs/1904.11489

Hardware:

GPU

Detector:

Public

Processing:

Batch

Last submitted:

November 04, 2019 (10 days ago)

Published:

November 04, 2019 at 04:46:19 CET

Submissions:

1

Open source:

No

Project page / code:

n/a

Reference:

J. Xu, Y. Cao, Z. Zhang, H. Hu. Spatial-Temporal Relation Networks for Multi-Object Tracking. In ICCV, 2019.

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
48.573.71.517.0 % 34.9 % 9,03884,1787472,919GPUPublic
IDF1ID PrecisionID Recall
53.972.842.8

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
MOT16-0136.749.471.20.92317.4 % 34.8 % 4053,60342139
MOT16-0355.657.973.92.714824.3 % 16.9 % 4,08542,0502701,449
MOT16-0649.657.672.60.422117.2 % 41.6 % 4435,31158280
MOT16-0745.950.472.91.85416.7 % 22.2 % 8767,835126347
MOT16-0833.542.776.71.56312.7 % 38.1 % 93210,11283223
MOT16-1242.658.175.00.98624.4 % 45.3 % 8243,91126110
MOT16-1429.839.272.22.01647.9 % 39.6 % 1,47311,356142371

Raw data:

n/a


STRN_MOT16