ST-HOI: A Spatial-Temporal Baseline for Human-Object Interaction Detection in Videos

Benchmark Model Rank Results
human-object-interaction-anticipation-onSTTRAN#2Person-wise Top5: t=1(mAP@0.5): 29.09
human-object-interaction-detection-on-vidhoiSTTRAN#2Detection: Full (mAP@0.5): 7.61Detection: Non-Rare (mAP@0.5): 13.18