Child Dangerous Climbing Action Recognition Dataset

The current security industry faces significant challenges in child safety monitoring, especially in home environments where children's climbing on hazardous objects often leads to accidental incidents. Existing monitoring systems often fail to timely recognize and warn of these dangerous behaviors, resulting in frequent accidents. This dataset aims to provide high-quality videos of children climbing to aid in the development of more accurate action recognition models, thereby enhancing household safety assurance. During data collection, high-resolution camera equipment was used in real home environments to ensure data authenticity and diversity. To ensure data quality, multiple rounds of annotation and expert review were conducted to guarantee consistency and accuracy of labeling. Data is stored in MP4 format for easy subsequent processing and analysis.