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


Short Name: tracks
Complete name: tracks
Description:

tracks

Runtime:

little

Submissions:

1

Benchmark
performance:

MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
-9.7983 -14.3934 17.9989 2.8332 60.1259 24508 48373 3621 80 3865 67.7544

Detailed
performance:


Video ID MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
20 9.0612 7.3836 26.5928 22.4490 46.9388 5091 5313 445 5 558 69.6890
21 0.3988 -3.6012 9.3023 0.0000 73.3333 77 879 43 1 35 64.4431
22 2.2458 -3.9289 12.5720 1.9417 47.5728 1301 7352 836 15 765 64.4496
23 -25.6690 -31.0297 17.2828 0.0000 42.8571 2517 2991 127 6 272 65.0162
24 -21.1749 -29.8187 6.0538 0.3040 75.6839 3783 11847 645 29 544 60.3977
25 -27.4374 -32.7032 14.4175 0.0000 56.8966 2590 4871 277 8 257 61.8148
26 -39.7043 -41.8543 5.7471 0.0000 88.2353 1803 3327 67 2 101 62.5022
27 -7.4592 -12.9710 26.7353 8.0645 30.6452 2497 2560 403 7 452 71.0780
28 10.4224 2.5266 24.3868 5.0000 35.0000 613 1439 217 5 210 68.6279
29 -4.6491 -6.4533 27.4015 4.0000 32.0000 829 1420 92 1 146 70.2673
30 -12.9602 -12.9602 15.5852 0.0000 35.5556 644 1466 130 0 132 65.8702
31 -77.6316 -77.6316 8.2880 4.4444 93.3333 1933 1925 57 0 54 71.4312
32 22.5898 21.8341 31.8737 10.0000 45.0000 830 2983 282 1 339 71.6083