Construction of Information Management Model for College Students Based on Deep Learning Algorithms and Data Collection

Abstract

In order to improve the efficiency of college student management and provide effective tools and assistants for college student management staff, the author proposes the construction of an information technology model for college student management based on deep learning algorithms and data collection. Firstly, use the FasterR-CNN model to detect the heads of personnel in the laboratory, Then, based on the output results of model detection, use the IoU algorithm to filter out duplicate detected targets, Finally, a coordinate based positioning method is used to determine whether there are people on each workbench in the laboratory, and the corresponding data is stored in the database. The main functions of this system include: (1) Real-time video monitoring and remote management of the laboratory, (2) Timed automatic photography detection and data collection provide data support for quantitative management in the laboratory, (3) Query and visualization of data on changes in laboratory personnel. The experimental findings demonstrate that our proposed model excels with an F1 Score exceeding 91%, showcasing robust generalization across detection confidence levels ranging from 50% to 99%. Notably, at a detection confidence of 96%, our model achieves its peak performance with an impressive F1 Score of 95.7%. This underscores the model’s exceptional detection capabilities. Leveraging Faster R-CNN and IoU optimization, our laboratory personnel statistics and management system offer real-time personnel tracking and remote management functionalities tailored for office environments.