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


Track-5: Pedestrian Detection in Complex and Crowded Events

Trackers with "(Baseline)" in the name are submitted by the HiEve team. Results are obtained by running their official code.

Team Name MR AP JI
first-result

0.7198 0.5890 0.4905
Anonymous submission
second-result

0.7168 0.5912 0.4932
Anonymous submission
2nd_submission

0.7147 0.5981 0.5026
Anonymous submission
result

0.7103 0.6072 0.5157
Y Tang, B Li, M Liu, et al.. AutoPedestrian: An Automatic Data Augmentation and Loss Function Search Scheme for Pedestrian Detection. In TIP, 2021
RetinaNet (Baseline)

0.7041 0.6082 0.4898
TY Lin, P Goyal, R Girshick, K He. Focal loss for dense object detection. In ICCV, 2017
FasterRCNN (Baseline)

0.6722 0.5935 0.5156
S Ren, K He, R Girshick, et al.. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. In NIPS, 2015