Benchmark:
MOT16 |
Short name:
CRF_TRACK
Detector:
Public
Description:
Reference:
Jun xiang, Chao Ma, Guohan Xu, Jianhua Hou, End-to-End Learning Deep CRF models for Multi-Object Tracking. In IEEE Transactions on Circuits and Systems for Video Technology, 2020
Last submitted:
June 13, 2019 (5 years ago)
Published:
June 13, 2019 at 09:46:50 CET
Submissions:
1
Project page / code:
n/a
Open source:
Yes
Hardware:
Runtime:
1.5 Hz
Benchmark performance:
Sequence | MOTA | IDF1 | HOTA | MT | ML | FP | FN | Rcll | Prcn | AssA | DetA | AssRe | AssPr | DetRe | DetPr | LocA | FAF | ID Sw. | Frag |
MOT16 | 50.3 | 54.4 | 40.7 | 139 (18.3) | 271 (35.7) | 7,148 | 82,746 | 54.6 | 93.3 | 41.6 | 40.1 | 44.2 | 74.6 | 42.5 | 72.7 | 78.5 | 1.2 | 702 (12.9) | 1,387 (25.4) |
Detailed performance:
Sequence | MOTA | IDF1 | HOTA | MT | ML | FP | FN | Rcll | Prcn | AssA | DetA | AssRe | AssPr | DetRe | DetPr | LocA | FAF | ID Sw. | Frag |
MOT16-01 | 47.9 | 62.5 | 47.8 | 10 | 6 | 464 | 2,854 | 55.4 | 88.4 | 57.1 | 40.2 | 61.5 | 74.0 | 43.3 | 69.1 | 76.6 | 1.0 | 12 | 45 |
MOT16-03 | 56.9 | 56.0 | 41.8 | 31 | 27 | 2,967 | 41,815 | 60.0 | 95.5 | 40.0 | 44.1 | 42.4 | 73.6 | 46.5 | 74.1 | 78.7 | 2.0 | 309 | 605 |
MOT16-06 | 48.7 | 60.7 | 45.8 | 43 | 100 | 438 | 5,413 | 53.1 | 93.3 | 52.6 | 40.1 | 56.1 | 77.6 | 42.4 | 74.6 | 79.8 | 0.4 | 63 | 128 |
MOT16-07 | 48.2 | 51.3 | 38.0 | 9 | 11 | 720 | 7,649 | 53.1 | 92.3 | 37.9 | 38.6 | 39.9 | 75.5 | 40.9 | 71.2 | 77.5 | 1.4 | 89 | 184 |
MOT16-08 | 36.6 | 48.5 | 37.6 | 16 | 21 | 1,237 | 9,290 | 44.5 | 85.8 | 43.4 | 32.7 | 46.1 | 77.3 | 35.6 | 68.7 | 79.2 | 2.0 | 89 | 153 |
MOT16-12 | 46.5 | 57.4 | 44.2 | 18 | 36 | 405 | 4,001 | 51.8 | 91.4 | 51.2 | 38.3 | 53.6 | 80.9 | 41.2 | 72.8 | 80.8 | 0.5 | 34 | 51 |
MOT16-14 | 31.0 | 43.3 | 29.6 | 12 | 70 | 917 | 11,724 | 36.6 | 88.1 | 33.0 | 26.6 | 35.1 | 70.0 | 28.0 | 67.5 | 75.7 | 1.2 | 106 | 221 |
Raw data: