Robust Online Multi-Object Tracking based on Tracklet Confidence and Online Discriminative Appearance Learning

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

Short name:

TC_ODAL

Benchmark:

Description:

Online multi-object tracking aims at producing complete tracks of multiple objects using the information accumulated up to the present moment. It still remains a difficult problem in complex scenes, because of frequent occlusion by clutter or other objects, similar appearances of different objects, and other factors. In this paper, we propose a robust online multi-object tracking method that can handle these difficulties effectively.

We first propose the tracklet confidence using the detectability and continuity of a tracklet, and formulate a multi-object tracking problem based on the tracklet confidence. The multi-object tracking problem is then solved by associating tracklets in different ways according to their confidence values. Based on this strategy, tracklets sequentially grow with online-provided detections, and fragmented tracklets are linked up with others without any iterative and expensive associations. Here, for reliable association between tracklets and detections, we also propose a novel online learning method using an incremental linear discriminant analysis for discriminating the appearances of objects. By exploiting the proposed learning method, tracklet association can be successfully achieved even under severe occlusion. Experiments with challenging public datasets show distinct performance improvement over other batch and online tracking methods.

[Parameters]
show_scan=4
new_thr=5
obs_thr=0.32508
type_thr=0.61126
scale_unit=2.827
std_height=180
pos_var=30, 75
alpha=0.44719
lambda=0.90809
atten=0.85705
init_prob=0.67935
tmplsize=64, 32
Bin=64
vecsize=2048
subregion=1
colortype=HSV
q=3.8414
ppp=4.153
use_ILDA=1
n_update=0
eigenThreshold=0.01
up_ratio=1
duration=5

Hardware:

2.6 GHz, 16 Cores

Detector:

Public

Processing:

Online

Last submitted:

February 20, 2015 (2 years ago)

Published:

November 01, 2014 at 03:09:33 CET

Submissions:

1

Open source:

Yes

Project page / code:

Reference:

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
15.170.52.23.2 % 55.8 % 12,97038,5386371,7162.6 GHz, 16 CoresPublic
IDF1ID PrecisionID Recall
0.00.00.0

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
TUD-Crossing55.80.072.80.51323.1 % 7.7 % 1103601735
PETS09-S2L230.20.069.22.5422.4 % 19.0 % 1,0745,375284499
ETH-Jelmoli31.20.072.01.14513.3 % 33.3 % 4951,2272384
ETH-Linthescher14.10.073.30.21971.0 % 78.7 % 2927,3522678
ETH-Crossing16.30.073.60.3260.0 % 73.1 % 6977016
AVG-TownCentre1.60.069.02.02260.0 % 71.7 % 8996,11126114
ADL-Rundle-1-0.20.069.99.03215.6 % 18.8 % 4,4764,78270286
ADL-Rundle-316.60.072.44.7446.8 % 22.7 % 2,9665,45066138
KITTI-1627.30.070.71.0170.0 % 17.6 % 21398637106
KITTI-1912.90.066.51.3624.8 % 27.4 % 1,3513,23766268
Venice-113.80.071.32.3170.0 % 35.3 % 1,0252,88821102

Raw data:


TC_ODAL
CEM