Tracktor++v2: Tracktor++ PyTorch 1.3

MOT20-06


Benchmark:

Short name:

Tracktor++v2

Detector:

Public

Description:

The problem of tracking multiple objects in a video sequence poses several challenging tasks. For tracking-by- detection these include object re-identification, motion prediction and dealing with occlusions. We present a tracker that accomplishes tracking without specifically targeting any of these tasks, in particular, we perform no training or optimization on tracking data. To this end, we exploit the bounding box regression of an object detector to predict the position of an object in the next frame, thereby converting a detector into a Tracktor. We demonstrate the extensibility of our Tracktor and provide a new state-of-the-art on three multi-object tracking benchmarks by extending it with a straightforward re-identification and camera motion compensation. This benchmark submission presents the results of our extended Tracktor++ multi-object tracker.

Reference:

P. Bergmann, T. Meinhardt, L. Leal-Taixé. Tracking without bells and whistles. In ICCV, 2019.

Last submitted:

November 30, 2020 (9 months ago)

Published:

November 30, 2020 at 16:50:13 CET

Submissions:

2

Open source:

Yes

Hardware:

Titan X

Runtime:

1.2 Hz

Benchmark performance:

Sequence MOTA IDF1 HOTA MT ML FP FN Rcll Prcn AssA DetA AssRe AssPr DetRe DetPr LocA FAF ID Sw. Frag
MOT2052.652.742.1365 (29.4)331 (26.7)6,930236,68054.397.642.042.345.971.644.179.382.41.51,648 (30.4)4,374 (80.6)

Detailed performance:

Sequence MOTA IDF1 HOTA MT ML FP FN Rcll Prcn AssA DetA AssRe AssPr DetRe DetPr LocA FAF ID Sw. Frag
MOT20-0472.765.451.5298552,85571,16474.098.646.257.550.473.160.380.382.61.47391,928
MOT20-0630.133.226.6281371,74590,50931.896.028.624.731.465.925.577.081.41.75121,408
MOT20-0750.149.639.3272325216,12751.398.538.440.442.472.442.180.883.50.4146304
MOT20-0821.027.223.4121162,07858,88024.090.029.518.732.066.919.472.981.12.6251734

Raw data:


Tracktor++v2