Automatic Crowd Counting on the Beaches of the Balearic Islands Using Convolutional Neural Networks

Résumé fourni par la source

This paper presents a system for real-time crowd estimation on beaches in the Balearic Islands, aimed at improving public access to information on coastal space usage. We consider two crowd counting models in the literature, that predict sparse locations of individuals and density maps, respectively. After fine-tuning both models using a custom dataset collected from various beaches across the Balearic Islands, we compare their performance in terms of total count. The resulting predictions are executed hourly and made available through a public website featuring an interactive map and time-series visualizations of beach occupancy.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.