Experimental Teaching Environment Student Operation Behavior Recognition Dataset

Currently, in the field of education during experimental teaching, it is difficult for teachers to monitor each student's operational behavior in real-time and accurately, posing a challenge to teaching quality and personalized guidance for students. Existing educational monitoring systems mostly rely on teachers' observations and manual records, which are inefficient and prone to omissions. This dataset aims to solve the problem of student behavior recognition in experimental teaching through video analysis technology, enhancing the intelligence level of teaching management. The data is collected using high-definition cameras in a real experimental teaching environment, ensuring good lighting and camera angles during collection. The data quality is ensured through multiple rounds of annotation and consistency checks, reviewed by educational technology experts. The annotation team includes 50 professionals with a teaching background. Data preprocessing includes video editing, keyframe extraction, noise removal, and other steps, stored in MP4 format, organized according to the type of operation. The overall data structure includes timestamps for each video segment, behavior labels, and related metadata information.