TeG: Temporal-Granularity Method for Anomaly Detection with Attention in Smart City Surveillance

TL;DR

This paper introduces TeG, a novel anomaly detection method for smart city surveillance that combines multi-scale spatio-temporal features using attention mechanisms, validated on an extended dataset in real-time city environments.

Contribution

The paper proposes a new temporal-granularity approach with attention mechanisms for anomaly detection, extending the UCF-Crime dataset with relevant new anomaly types.

Findings

Achieved successful real-time anomaly detection in city surveillance systems.

Enhanced detection accuracy with multi-scale spatio-temporal features.

Validated effectiveness on an extended, real-world dataset.

Abstract

Anomaly detection in video surveillance has recently gained interest from the research community. Temporal duration of anomalies vary within video streams, leading to complications in learning the temporal dynamics of specific events. This paper presents a temporal-granularity method for an anomaly detection model (TeG) in real-world surveillance, combining spatio-temporal features at different time-scales. The TeG model employs multi-head cross-attention blocks and multi-head self-attention blocks for this purpose. Additionally, we extend the UCF-Crime dataset with new anomaly types relevant to Smart City research project. The TeG model is deployed and validated in a city surveillance system, achieving successful real-time results in industrial settings.