AP_HWDPL_p: Appearance Model with R-CNN

TUD-Crossing


Benchmark:

Short name:

AP_HWDPL_p

Detector:

Public

Description:

Online tracking with appearance model based on R-CNN from Huawei (HWDPL).

Reference:

C. Long, A. Haizhou, S. Chong, Z. Zijie, B. Bo. Online Multi-Object Tracking with Convolutional Neural Networks. In 2017 IEEE International Conference on Image Processing (ICIP), 2017.

Last submitted:

December 13, 2016 (3 years ago)

Published:

September 26, 2017 at 09:20:18 CET

Submissions:

2

Project page / code:

n/a

Open source:

No

Hardware:

GTX 1080, 2.3 GHz, 1 Core

Runtime:

6.7 Hz

Benchmark performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
2D MOT 201538.547.172.663 (8.7)270 (37.4)4,00533,20346.087.60.7586 (12.8)1,263 (27.5)

Detailed performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
ADL-Rundle-135.343.672.24105735,41341.887.21.13981
ADL-Rundle-343.351.977.6796435,07650.188.81.04785
AVG-TownCentre28.444.766.99639414,00544.077.02.1169412
ETH-Crossing34.242.379.51162063037.294.90.11013
ETH-Jelmoli52.965.575.910131781,00160.589.60.41637
ETH-Linthescher39.447.575.2171241855,18541.995.30.24471
KITTI-1640.761.168.93216083151.184.50.81835
KITTI-1934.350.667.72206632,80547.579.30.643147
PETS09-S2L238.934.370.8145525,16446.489.01.3179328
TUD-Crossing61.364.173.1521440163.698.00.11227
Venice-139.148.771.847762,69241.096.10.2927

Raw data: