Appearance Model with R-CNN

TUD-Crossing PETS09-S2L2 ETH-Jelmoli ETH-Linthescher ETH-Crossing AVG-TownCentre ADL-Rundle-1 ADL-Rundle-3 KITTI-16 KITTI-19

Short name:

AP_HWDPL

Benchmark:

Description:

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

Hardware:

GTX 1080, 2.3 GHz, 1 Core

Detector:

Private

Processing:

Online

Last submitted:

December 10, 2016 (2 years ago)

Published:

September 26, 2017 at 09:16:42 CET

Submissions:

3

Open source:

No

Project page / code:

n/a

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.

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
53.075.50.929.1 % 20.2 % 5,15922,9847081,476GTX 1080, 2.3 GHz, 1 CorePrivate
IDF1ID PrecisionID Recall
52.262.844.6

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
TUD-Crossing77.958.875.80.01353.8 % 0.0 % 22251617
PETS09-S2L256.037.974.00.94214.3 % 4.8 % 3873,636216368
ETH-Jelmoli61.267.377.70.34526.7 % 26.7 % 1218432132
ETH-Linthescher61.255.077.90.319723.4 % 36.5 % 3313,03899166
ETH-Crossing61.164.481.90.22619.2 % 34.6 % 54329722
AVG-TownCentre64.069.471.71.022639.8 % 12.4 % 4651,963147364
ADL-Rundle-145.151.274.92.93231.3 % 15.6 % 1,4553,60255136
ADL-Rundle-346.849.381.90.94418.2 % 18.2 % 5664,7954381
KITTI-1652.469.474.11.11735.3 % 11.8 % 2305621737
KITTI-1944.356.068.40.96221.0 % 9.7 % 9062,01060199
Venice-141.931.476.41.41741.2 % 11.8 % 6421,9812754

Raw data:


AP_HWDPL