Challenge and Dataset on Large-scale Human-centric Video Analysis in Complex Events (HiEve)

Short Name: ByteTrack
Complete name: mot by associating every detection box
Description:

n/a

Runtime:

30.5

Submissions:

7

Benchmark
performance:

MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
61.6720 56.4451 63.1294 38.3001 21.6159 2822 22852 1031 91 1266 78.0294

Detailed
performance:

Video ID MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
20 86.0939 84.0808 82.9697 75.5102 6.1224 140 1493 26 6 51 79.7662
21 30.8076 26.8076 30.2835 10.0000 30.0000 106 560 28 1 34 69.9539
22 38.6319 32.8688 32.2091 24.2718 15.5340 823 4701 433 14 482 72.8615
23 91.0794 86.6121 79.8686 78.5714 7.1429 114 255 31 5 35 81.5248
24 21.2866 9.6622 29.5329 9.4225 47.1125 431 9909 232 39 244 73.9449
25 74.8024 69.5366 85.0359 50.5747 0.0000 186 1337 7 8 37 76.8827
26 62.6344 60.4844 48.6247 41.1765 23.5294 220 1116 54 2 64 78.0540
27 88.1716 82.6598 79.9393 87.0968 3.2258 141 418 42 7 38 81.4744
28 64.0742 56.1784 64.3290 55.0000 10.0000 304 579 27 5 65 73.2706
29 70.5856 68.7814 65.3494 68.0000 12.0000 149 473 36 1 54 80.1155
30 68.8855 68.8855 73.3711 48.8889 0.0000 74 510 33 0 46 78.2590
31 90.9256 90.9256 92.8022 93.3333 2.2222 41 156 3 0 17 81.6728
32 71.3233 69.0562 66.4022 42.5000 25.0000 93 1345 79 3 99 77.5335