Low-Latency Video Anonymization for Crowd Anomaly Detection: Privacy Versus Performance

TL;DR

This paper introduces LA3D, a lightweight adaptive anonymization method that enhances privacy protection in real-time video anomaly detection without significantly compromising detection performance.

Contribution

It proposes a novel, computationally efficient AN approach that balances privacy and utility, outperforming existing methods in real-time surveillance scenarios.

Findings

LA3D significantly improves privacy anonymization.

LA3D maintains high anomaly detection utility.

It outperforms conventional and deep learning approaches.

Abstract

Recent advancements in artificial intelligence hold ample potential for monitoring applications using surveillance cameras. However, concerns about privacy and model bias have made it challenging to utilize them in public. Although de-identification approaches have been proposed in the literature, aiming to achieve a certain level of anonymization (AN), most of them employ deep learning models that are computationally demanding for real-time edge deployment. This study revisits conventional AN solutions for privacy protection and real-time video anomaly detection (VAD) applications. We propose a lightweight adaptive AN for VAD (LA3D) that employs dynamic adjustment to enhance full-body privacy protection. We have evaluated privacy protection and VAD utility retention efficacy using several publicly available datasets to examine the strengths and weaknesses of different AN methods and highlight the promising leverage of our approach. Our experiment demonstrates that the LA3D enables substantial improvement in privacy AN without severely degrading VAD efficacy, outperforming conventional and deep learning approaches. Code is available at https://github.com/muleina/LA3D .