Unsupervised Abnormal Crowd Activity Detection Using Semiparametric Scan Statistic

Detection and Localization of Anomalous Motion in Video Sequences from Local Histograms of Labeled Affine Flows

“…Other works are able to detect anomalies locally in videos and without an explicit definition of what the abnormality is. Among these, two main classes are found: trajectory-based (Stauffer and Grimson, 2000;Piciarelli et al, 2008;Wu et al, 2010;Jiang et al, 2011;Zen et al, 2012;Li et al, 2013) and feature-based ones (Adam et al, 2008;Kim and Grauman, 2009;Kratz and Nishino, 2009;Antic and Ommer, 2011;Bertini et al, 2012;Cong et al, 2013;Hu et al, 2013;Li et al, 2014;Cheng et al, 2015;Zhang et al, 2016).…”

“…The method presented in Hu et al (2013) exploits optic flow measurements only and is fully unsupervised. It introduces a semiparametric likelihood test computed on a given window and outside the window to decide if the content of the tested window contains abnormal motion or not.…”

“…We report comparative results with the following motion-based anomaly detection methods: the method based on sparse reconstruction error (SRC) (Cong et al, 2013) which exploits multiscale histograms of optical flow, the method relying on chaotics invariants (CI) (Wu et al, 2010), the method involving the social force model (SF) (Mehran et al, 2009), the method built upon scan statistic (SS) (Hu et al, 2013), and the method introducing motion influence maps (MIM) (Lee et al, 2015). Sample visual results are gathered in Figure 6.…”

“…It may depend on the context and the application. As in Chandola et al (2009) and Hu et al (2013), we consider in this work that anomalous motion means that the motion significantly differs from the mainstream one, observed either in the same video segment or in the whole video. Indeed, anomalous motion is taken here in the broad acceptance of a different behavior w.r.t.…”

A Compilation of Methods and Datasets for Group and Crowd Action Recognition

“…It is important to highlight the work of Wu et al (Wu et al, 2015). They proposed the MoSIF method is combined with HMM (Wu et al-, 2015) A general trend could be noticed and described as the actions considered normal ones have an extended dumtion -.. a general trend that would be described as that the behaviors considered normal ones have an extended du¡ation, in which most people make up the crowd, while the behaviors considered abnormal are caused by few people in the croud and in shof times of du¡ation-For the study of these types of behaviors, Hu et al (Hu et al-, 2013) proposed to use a statistical exploration method analyzing the video in a separate way as sliding windows in which the behavion considered anomalous are detected, taking into account that the algorithm used in this technique requires monitoring.…”