FGRNetIV: Flow-guided Geometric Regression Network with Identity Verification

TUD-Crossing


Benchmark:

Short name:

FGRNetIV

Detector:

Public

Description:

Reference:

Last submitted:

December 15, 2020 (2 months ago)

Published:

December 15, 2020 at 12:41:19 CET

Submissions:

1

Project page / code:

n/a

Open source:

No

Hardware:

2080Ti

Runtime:

1.4 Hz

Benchmark performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
MOT1547.547.876.1169 (23.4)190 (26.4)5,53125,50258.586.71.01,240 (21.2)1,550 (26.5)

Detailed performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
ADL-Rundle-138.850.774.21142,1033,53362.073.34.260162
ADL-Rundle-347.648.481.7875374,72353.591.00.96879
AVG-TownCentre47.146.371.170359162,37066.883.92.0498453
ETH-Crossing43.449.583.83103151948.394.00.11817
ETH-Jelmoli57.763.980.5151330174170.885.60.73244
ETH-Linthescher50.355.380.732962174,16153.495.60.25989
KITTI-1651.355.270.53110169359.390.90.53454
KITTI-1946.753.071.810174572,24458.087.10.4146145
PETS09-S2L248.829.874.9242834,38354.594.90.6274438
TUD-Crossing77.653.474.8901521380.798.30.11925
Venice-144.744.775.4635701,92257.982.21.33244

Raw data: