Identification of Unusual Events for Smart City Surveillance—A Federated Approach
TL;DR
This article proposes a federated learning approach for identifying unusual events in smart city surveillance systems, addressing data privacy by training AI models locally without sharing raw video data.
Identification of Unusual Events for Smart City Surveillance—A Federated Approach
Tehreem Qamar; Narmeen Zakaria Bawany; Muhammad Hussain Mughal
https://doi.org/10.1109/ACCESS.2025.3648440
Volume 14
With the advent of innovative technologies, smart city video surveillance systems are prevailing increasingly. These systems collect extensive amounts of data using cameras and sensors strategically placed throughout the city. They typically utilize artificial intelligence models that are developed on a classical machine learning approach in which data is collected and shared for model development...