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


Short Name: DeepBlueAI
Complete name: DeepBlueAI
Description: Tired
Runtime: 20min Hz
Submissions: 1

Benchmark
performance:

MOTA MOTP Total_AP
55.1543 52.3774 65.9403

Detailed
performance:


Video Name MOTA MOTP Total_AP
hm_in_waiting_hall (ID:20) 54.4557 55.8061 80.7472
hm_in_bus (ID:21) 31.7244 32.0742 34.7395
hm_in_dining_room2 (ID:22) -10.6723 8.1647 31.6295
hm_in_lab2 (ID:23) 67.9720 62.6539 81.6848
hm_in_subway_station (ID:24) 36.5328 39.4063 47.7066
hm_in_passage (ID:25) 61.7239 72.5388 69.2177
hm_in_fighting4 (ID:26) 71.5209 42.5622 68.4857
hm_in_shopping_mall3 (ID:27) 62.1769 58.6635 76.3027
hm_in_restaurant (ID:28) 58.1410 54.7813 56.9331
hm_in_accident (ID:29) 66.4479 48.2173 73.1205
hm_in_stair3 (ID:30) 53.6025 25.3416 58.7849
hm_in_crossroad (ID:31) 75.9008 70.0137 82.4194
hm_in_robbery (ID:32) 44.0829 45.2804 50.3821