Method

Phase Enhance SORT [PESORT]


Submitted on 17 Oct. 2023 05:18 by
zheng yujin (None)

Running time:0.04 s
Environment:GPU @ 2.0 Ghz (Python)

Method Description:
By calculating the offset of two adjacent frames in
the video sequence by the phase correlation method,
and matching the detection frame with the same
displacement, the impact caused by camera motion can
be effectively reduced
Parameters:
conf=0.01
nms=0.7
track_thresh=0.6
track_buffer=30
match_thresh=0.9
Latex Bibtex:

Detailed Results

From all 29 test sequences, our benchmark computes the commonly used tracking metrics CLEARMOT, MT/PT/ML, identity switches, and fragmentations [1,2]. The tables below show all of these metrics.


Benchmark MOTA MOTP MODA MODP
CAR 28.86 % 83.96 % 28.91 % 90.04 %
PEDESTRIAN 44.19 % 76.23 % 44.72 % 93.21 %

Benchmark recall precision F1 TP FP FN FAR #objects #trajectories
CAR 37.00 % 89.92 % 52.43 % 13472 1511 22939 13.58 % 22756 401
PEDESTRIAN 53.49 % 86.71 % 66.17 % 12515 1918 10880 17.24 % 20995 263

Benchmark MT PT ML IDS FRAG
CAR 22.62 % 17.38 % 60.00 % 16 58
PEDESTRIAN 24.40 % 37.11 % 38.49 % 121 535

This table as LaTeX

[1] K. Bernardin, R. Stiefelhagen: Evaluating Multiple Object Tracking Performance: The CLEAR MOT Metrics. JIVP 2008.
[2] Y. Li, C. Huang, R. Nevatia: Learning to associate: HybridBoosted multi-target tracker for crowded scene. CVPR 2009.


eXTReMe Tracker