CNNTCM:

TUD-Crossing


Benchmark:

Short name:

CNNTCM

Detector:

Public

Description:

n/a

Reference:

B. Wang, K. L. Chan, L. Wang, B. Shuai, Z. Zuo, T. Liu, G. Wang. Joint Learning of Convolutional Neural Networks and Temporally Constrained Metrics for Tracklet Association. In DeepVision Workshop (CVPR), 2016.

Last submitted:

September 27, 2015 (4 years ago)

Published:

October 11, 2015 at 05:27:01 CET

Submissions:

1

Project page / code:

n/a

Open source:

No

Hardware:

2.60GHz, 4 core

Runtime:

1.7 Hz

Benchmark performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
2D MOT 201529.636.871.881 (11.2)317 (44.0)7,78634,73343.577.41.3712 (16.4)943 (21.7)

Detailed performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
ADL-Rundle-118.524.173.06112,3035,14944.764.34.6128121
ADL-Rundle-327.835.873.44101,7745,46546.272.62.810391
AVG-TownCentre31.345.969.538759523,82546.577.72.1137246
ETH-Crossing26.641.775.11181072527.796.50.012
ETH-Jelmoli41.352.174.58132331,22651.784.90.52938
ETH-Linthescher14.623.576.58158727,54615.595.10.1611
KITTI-1648.060.772.90313373656.787.90.61634
KITTI-1931.147.865.95148812,73248.974.80.871155
PETS09-S2L249.633.970.7557803,88659.788.11.8192218
TUD-Crossing60.568.373.7636635268.191.90.31714
Venice-119.238.074.1075823,09132.371.71.31213

Raw data: