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

Short Name: Refine FairMOT
Complete name: Refine FairMOT
Description:

Refine FairMOT

Runtime:

na

Submissions:

6

Benchmark
performance:

MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
43.5551 38.2708 48.2311 21.3012 32.4239 3155 34612 1561 92 2756 76.5541

Detailed
performance:

Video ID MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
20 65.8340 63.8209 68.2720 40.8163 12.2449 327 3672 77 6 362 79.2983
21 25.0249 21.0249 37.1297 0.0000 26.6667 107 594 51 1 62 67.0499
22 17.1011 10.5147 19.1879 3.8835 32.0388 726 6767 554 16 642 69.2210
23 81.9358 76.5751 61.7538 64.2857 21.4286 71 680 59 6 107 81.5660
24 9.9620 -1.6624 17.6830 2.4316 63.5258 600 11243 250 39 345 69.6956
25 65.5138 60.2480 78.5812 39.6552 0.0000 441 1590 63 8 159 77.6421
26 44.4355 43.3605 38.1532 29.4118 23.5294 240 1767 60 1 137 77.3994
27 77.6816 72.1698 56.8944 66.1290 3.2258 166 784 184 7 295 78.3273
28 15.5152 6.0403 17.0579 0.0000 65.0000 30 2071 39 6 107 69.0905
29 29.9508 28.1466 32.8657 0.0000 48.0000 28 1508 31 1 134 75.8856
30 57.0852 57.0852 64.7148 26.6667 2.2222 65 718 68 0 104 77.6132
31 69.1016 69.1016 71.3123 57.7778 6.6667 86 562 33 0 123 70.4274
32 42.9868 42.2311 50.0244 22.5000 37.5000 268 2656 92 1 179 75.5047