MPTC: Multi-Pedestrian Tracking with Clusters

MOT20-04


Benchmark:

MOT17 | MOT20 | MOT16 |

Short name:

MPTC

Detector:

Public

Description:

Reference:

D. Stadler, J. Beyerer. Multi-Pedestrian Tracking with Clusters. In 2021 17th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), 2021.

Last submitted:

December 22, 2020 (1 year ago)

Published:

January 06, 2022 at 09:13:25 CET

Submissions:

1

Project page / code:

n/a

Open source:

No

Hardware:

Nvidia Tesla V100

Runtime:

0.7 Hz

Benchmark performance:

Sequence MOTA IDF1 HOTA MT ML FP FN Rcll Prcn AssA DetA AssRe AssPr DetRe DetPr LocA FAF ID Sw. Frag
MOT2060.659.748.5635 (51.1)208 (16.7)45,318153,97870.288.946.550.751.666.456.872.081.410.14,533 (64.5)5,163 (73.5)

Detailed performance:

Sequence MOTA IDF1 HOTA MT ML FP FN Rcll Prcn AssA DetA AssRe AssPr DetRe DetPr LocA FAF ID Sw. Frag
MOT20-0478.774.359.44763720,22836,79686.692.155.263.960.572.371.075.682.69.71,4331,691
MOT20-0639.938.531.0718613,55564,24351.683.526.936.031.053.140.565.678.913.41,9342,090
MOT20-0770.656.046.86371,2268,15875.495.339.456.345.459.760.776.881.72.1333330
MOT20-0827.835.327.4257810,30944,78142.276.026.528.630.553.032.859.277.312.88331,052

Raw data: