Human in Events

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


Short Name: SGT_noExtraData
Complete name: Sparse Graph Tracker trained on only Hieve dataset
Description: Sparse Graph Tracker trained on only Hieve dataset
Runtime: 13.4 Hz
Submissions: 2

Benchmark
performance:

MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
47.2020 41.9751 53.6595 24.0294 28.7513 4699 30727 1361 91 2299 75.0732

Detailed
performance:


Video Name MOTA w_MOTA IDF1 MT ML FP FN ID_Sw ID_Sw_DT Frag MOTP
hm_in_waiting_hall (ID:20) 76.7477 74.3991 75.4893 79.5918 8.1633 895 1814 65 7 219 76.8991
hm_in_bus (ID:21) 11.9641 7.9641 16.8209 0.0000 63.3333 22 839 22 1 30 65.9961
hm_in_dining_room2 (ID:22) 16.4314 10.6683 21.6862 7.7670 46.6019 414 7433 265 14 464 70.6031
hm_in_lab2 (ID:23) 85.5263 81.0590 63.5497 78.5714 7.1429 169 432 48 5 65 78.3559
hm_in_subway_station (ID:24) 10.7736 -2.3412 23.1214 4.5593 50.1520 1265 10308 411 44 517 67.0532
hm_in_passage (ID:25) 54.1996 50.2503 67.1086 14.3678 1.1494 376 2262 143 6 261 72.3085
hm_in_fighting4 (ID:26) 47.5000 45.3500 49.3761 35.2941 29.4118 240 1669 44 2 113 75.4404
hm_in_shopping_mall3 (ID:27) 81.5981 77.6611 75.1937 85.4839 1.6129 309 534 92 5 143 79.4511
hm_in_restaurant (ID:28) 35.0178 27.1220 39.4844 10.0000 35.0000 358 1235 53 5 108 68.8580
hm_in_accident (ID:29) 43.9875 42.1833 45.0133 40.0000 20.0000 237 961 55 1 92 74.9609
hm_in_stair3 (ID:30) 56.3792 56.3792 63.0441 24.4444 2.2222 36 763 66 0 88 78.2852
hm_in_crossroad (ID:31) 82.1234 82.1234 84.1270 91.1111 2.2222 186 184 24 0 50 76.0878
hm_in_robbery (ID:32) 51.6446 50.8889 53.4497 20.0000 37.5000 192 2293 73 1 149 74.0877