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

Short Name: TCFM
Complete name: TCFM
Description:

TCFM

Runtime:

1

Submissions:

2

Benchmark
performance:

MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
43.9713 38.3998 45.1835 21.3012 29.5908 3821 33423 1794 97 2782 77.6975

Detailed
performance:

Video ID MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
20 74.3336 72.3205 64.4624 67.3469 8.1633 300 2656 106 6 330 80.8980
21 9.8704 5.8704 25.9358 0.0000 33.3333 162 672 70 1 86 67.1361
22 16.5654 10.8023 18.1848 5.8252 50.4854 306 7479 314 14 472 73.2898
23 73.2605 67.8998 49.3000 57.1429 14.2857 89 1057 53 6 102 81.8426
24 12.1808 -0.9340 19.5443 2.7356 56.2310 548 10863 384 44 488 69.9257
25 52.5692 47.9617 64.6724 20.6897 1.1494 531 2145 204 7 209 76.7482
26 34.8118 32.6618 36.0165 17.6471 23.5294 491 1856 78 2 187 75.5236
27 71.1080 65.5962 55.4219 66.1290 3.2258 312 994 162 7 311 80.3016
28 33.4386 25.5428 39.3107 20.0000 10.0000 399 1171 116 5 199 72.1036
29 39.2043 37.4001 40.7061 24.0000 16.0000 329 951 80 1 152 74.8498
30 47.2012 47.2012 56.7439 24.4444 4.4444 95 910 42 0 49 78.8648
31 81.0345 81.0345 66.7147 82.2222 4.4444 59 303 56 0 44 79.0237
32 49.0548 46.0319 45.8759 22.5000 27.5000 200 2366 129 4 153 75.6569