Dataset for Detecting Illegal Standing Under Lifting Operations

In industrial sites, illegal behavior by personnel often results in serious safety hazards. Current monitoring systems struggle to identify and respond to these dangerous behaviors in real-time, leading to frequent accidents. This dataset aims to solve this issue by providing rich video data of illegal standing behavior under lifting operations, aiding the development of more effective monitoring and early warning systems. Data collection was conducted using high-definition cameras in real construction site environments, focusing on illegal behavior under the paths of lifting objects. To ensure data quality, we implemented multiple rounds of annotation and expert reviews to guarantee data accuracy and consistency. All videos are saved in MP4 format and organized by timestamps to facilitate subsequent analysis and training.