This paper introduces Object-based Set Similarity (OSS), a novel metric designed to address computational costs and unreliable evaluations in active learning for real-world object detection. OSS quantifies active learning method effectiveness without requiring extensive detector training and improves evaluation robustness by identifying representative validation sets.

This study evaluates the performance of YOLOv7, an advanced object detection model, in identifying kitchen safety risks related to knife handling, specifically improper finger placement and blade contact with the hand. The findings highlight YOLOv7's strong potential for accurately detecting these hazards to improve kitchen safety.