Deep Crowd Anomaly Detection: State-of-the-Art, Challenges, and Future Research Directions

TL;DR

This paper reviews recent deep learning methods for crowd anomaly detection in smart cities, analyzing datasets, algorithms, and performance, and discusses future research challenges and directions.

Contribution

It provides a comprehensive taxonomy and performance comparison of recent algorithms, highlighting the limited impact of pre-trained model heterogeneity.

Findings

Pre-trained model heterogeneity has minimal effect on detection performance.

Deep learning methods generally outperform traditional machine learning approaches.

The paper identifies key challenges and promising future research directions.

Abstract

Crowd anomaly detection is one of the most popular topics in computer vision in the context of smart cities. A plethora of deep learning methods have been proposed that generally outperform other machine learning solutions. Our review primarily discusses algorithms that were published in mainstream conferences and journals between 2020 and 2022. We present datasets that are typically used for benchmarking, produce a taxonomy of the developed algorithms, and discuss and compare their performances. Our main findings are that the heterogeneities of pre-trained convolutional models have a negligible impact on crowd video anomaly detection performance. We conclude our discussion with fruitful directions for future research.