DSFC: Deep Sort Optical Flow Compensation with YOLOv4

TUD-Crossing


Benchmark:

MOT16 | MOT15 |

Short name:

DSFC

Detector:

Private

Description:

we design a motion flow and compensation strategy for predicting the location of occluded objects. Since the occluded objects may be legible in earlier frames, we utilize the velocity and location of the objects in the past frames to predict the possible location of the occluded objects. In addition, to improve the tracking speed and further enhance the tracking robustness, we utilize efficient YOLOv4, train on humanCrowd dateset to produce the detections in the proposed algorithm. By using YOLOv4, the tracking speed of our proposed method improved significantly.

Reference:

Last submitted:

November 16, 2021 (2 months ago)

Published:

November 19, 2021 at 11:52:07 CET

Submissions:

1

Project page / code:

n/a

Open source:

No

Hardware:

2.4 GHZ, 1 Core

Runtime:

10.8 Hz

Benchmark performance:

Sequence MOTA IDF1 HOTA MT ML FP FN Rcll Prcn AssA DetA AssRe AssPr DetRe DetPr LocA FAF ID Sw. Frag
MOT1555.159.545.7228 (31.6)166 (23.0)5,78621,30165.387.445.146.749.973.052.470.280.51.0504 (0.0)1,671 (0.0)

Detailed performance:

Sequence MOTA IDF1 HOTA MT ML FP FN Rcll Prcn AssA DetA AssRe AssPr DetRe DetPr LocA FAF ID Sw. Frag
ADL-Rundle-143.556.143.21131,8093,40063.576.645.941.150.273.950.160.579.53.649440
ADL-Rundle-347.053.042.51086984,63354.488.842.542.545.080.146.375.684.51.157147
AVG-TownCentre60.372.849.986276362,12570.388.852.647.858.271.353.067.076.51.476352
ETH-Crossing69.175.156.2461728671.597.755.856.758.781.259.681.483.70.1732
ETH-Jelmoli60.468.754.4191134764074.884.554.754.361.378.263.471.784.40.81740
ETH-Linthescher53.456.946.042952963,81657.394.546.345.651.377.848.379.884.20.25284
KITTI-1658.672.350.67116452569.187.853.248.357.270.353.567.976.80.81529
KITTI-1953.263.845.616107401,72867.783.046.445.151.570.151.663.375.60.735133
PETS09-S2L264.848.738.51819732,23576.888.427.753.632.856.660.569.678.92.2183286
TUD-Crossing82.880.557.990418083.799.654.262.061.571.964.977.279.70.0637
Venice-159.664.250.0641021,73362.096.551.948.358.175.251.079.383.10.2791

Raw data: