Entrance Repeated Glancing Action Recognition Dataset

The current security industry faces increasingly severe safety challenges, especially as the demand for surveillance in public places continues to grow. Existing video surveillance systems often rely on simple motion detection, which cannot effectively identify complex suspicious behaviors such as repeated glancing. This dataset aims to help researchers and developers build more intelligent action recognition systems, thereby improving the accuracy and effectiveness of public safety monitoring by providing high-quality video data. Data collection is conducted using high-resolution cameras in different environments to ensure a diverse range of monitoring scenarios. Data quality control combines multiple rounds of annotation and expert review to ensure consistency and accuracy of annotations. The data storage format is MP4, and all video files are organized by time and location.