Plug-and-Play CNN for Crowd Motion Analysis: An Application in Abnormal Event Detection

TL;DR

This paper introduces a novel plug-and-play CNN-based approach that combines semantic features with optical flow to detect local crowd abnormalities in videos without the need for fine-tuning, outperforming existing methods.

Contribution

It proposes a measure-based method leveraging CNN features and optical flow for abnormal event detection, avoiding fine-tuning and enhancing detection accuracy.

Findings

Outperforms state-of-the-art methods on challenging datasets

Effective in capturing local abnormalities without fine-tuning

Combines semantic CNN features with optical flow for improved detection

Abstract

Most of the crowd abnormal event detection methods rely on complex hand-crafted features to represent the crowd motion and appearance. Convolutional Neural Networks (CNN) have shown to be a powerful tool with excellent representational capacities, which can leverage the need for hand-crafted features. In this paper, we show that keeping track of the changes in the CNN feature across time can facilitate capturing the local abnormality. We specifically propose a novel measure-based method which allows measuring the local abnormality in a video by combining semantic information (inherited from existing CNN models) with low-level Optical-Flow. One of the advantage of this method is that it can be used without the fine-tuning costs. The proposed method is validated on challenging abnormality detection datasets and the results show the superiority of our method compared to the state-of-the-art methods.