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

Short Name: IQHAT_v1
Complete name: Identity-Quantity Harmonic Multi-Object Tracking
Description:

n/a

Runtime:

14

Submissions:

3

Benchmark
performance:

MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
53.4166 48.7641 57.4081 28.7513 28.7513 4430 27016 1011 81 1639 77.1049

Detailed
performance:

Video ID MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
20 73.0847 72.4137 80.3148 65.3061 8.1633 1351 1833 27 2 142 78.4779
21 15.2542 11.2542 31.6035 0.0000 26.6667 242 533 75 1 62 66.1738
22 34.4803 29.1289 33.7828 18.4466 33.0097 646 5449 265 13 284 72.3027
23 81.8912 76.5305 55.8994 71.4286 14.2857 105 640 67 6 128 80.7410
24 16.5810 5.8507 25.6811 3.9514 59.2705 291 10783 130 36 226 73.2069
25 54.6113 50.0038 71.3115 30.4598 0.0000 849 1769 138 7 172 73.2221
26 60.3495 58.1995 44.5594 35.2941 23.5294 86 1341 48 2 103 79.0237
27 82.5034 79.3538 74.5078 82.2581 6.4516 214 624 51 4 135 81.2809
28 57.3233 46.2692 50.1410 20.0000 40.0000 117 927 37 7 79 74.8618
29 67.3670 65.5628 59.5918 68.0000 12.0000 68 622 40 1 84 79.6657
30 59.2032 59.2032 62.8824 33.3333 2.2222 102 668 39 0 39 77.5117
31 89.5644 89.5644 92.6612 93.3333 2.2222 40 183 7 0 33 80.8125
32 61.2476 59.7362 56.9854 30.0000 25.0000 319 1644 87 2 152 75.1955