Deep Learning for Campus Pedestrian Volume Prediction: A Two-Stage Approach to Enhancing Information Security

Today's article comes from the IET journal of Information Security. The authors are Zhuang et al., from Fudan University, in China. In this paper, they showcase a system for predicting foot-traffic on a college campus. First, they transform the starting data into volume counts across a grid of campus regions. Then they train a transformer-based time-series model to forecast how those counts will change over time.

DOI: 10.1049/ise2/5895979