Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Sep 24;19(19):4145.
doi: 10.3390/s19194145.

An Efficient and Robust Unsupervised Anomaly Detection Method Using Ensemble Random Projection in Surveillance Videos

Jingtao Hu  1 , En Zhu  2 , Siqi Wang  3 , Xinwang Liu  4 , Xifeng Guo  5 , Jianping Yin  6
Affiliations

An Efficient and Robust Unsupervised Anomaly Detection Method Using Ensemble Random Projection in Surveillance Videos

Jingtao Hu et al. Sensors (Basel). .

Abstract

Video anomaly detection is widely applied in modern society, which is achieved by sensors such as surveillance cameras. This paper learns anomalies by exploiting videos under the fully unsupervised setting. To avoid massive computation caused by back-prorogation in existing methods, we propose a novel efficient three-stage unsupervised anomaly detection method. In the first stage, we adopt random projection instead of autoencoder or its variants in previous works. Then we formulate the optimization goal as a least-square regression problem which has a closed-form solution, leading to less computational cost. The discriminative reconstruction losses of normal and abnormal events encourage us to roughly estimate normality that can be further sifted in the second stage with one-class support vector machine. In the third stage, to eliminate the instability caused by random parameter initializations, ensemble technology is performed to combine multiple anomaly detectors' scores. To the best of our knowledge, it is the first time that unsupervised ensemble technology is introduced to video anomaly detection research. As demonstrated by the experimental results on several video anomaly detection benchmark datasets, our algorithm robustly surpasses the recent unsupervised methods and performs even better than some supervised approaches. In addition, we achieve comparable performance contrast with the state-of-the-art unsupervised method with much less running time, indicating the effectiveness, efficiency, and robustness of our proposed approach.

Keywords: random projection; surveillance camera; unsupervised ensemble learning; video anomaly detection.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
The workflow of our proposed method: (A) Preprocessing: Rescale and partition unlabeled videos into spatio-temporal cubes and extract features from them. (B) Normalcy Estimation: Roughly evaluate normality from all spatio-temporal cubes without any prior information. (C) Model Refinement: Eliminate remaining abnormality and build a model of normalcy. (D) Inference Ensemble: Infer the anomaly scores of all video events through the normalcy model under different settings and ensemble the scores.
Figure 2
Figure 2
An example of reconstruction discriminative distribution of UCSD Ped2 Dataset.
Figure 3
Figure 3
The schematic diagram of RR-Net structure.
Figure 4
Figure 4
An example of one-class SVM building a tight boundary on a synthetic data with a peanut shape distribution.
Figure 5
Figure 5
An illustration of our proposed anomaly scores ensemble under different initializations and final anomaly scores computation.
Figure 6
Figure 6
The representative abnormal events in the benchmark datasets (highlighted in red box).
Figure 7
Figure 7
The visualization results of our method on UCSD Ped1 dataset (highlighted in red).
Figure 8
Figure 8
The visualization results of our method on UCSD Ped2 dataset (highlighted in red).
Figure 9
Figure 9
The visualization results of our method detection on Avenue dataset (highlighted in red).
Figure 10
Figure 10
The EER and AUC comparison (mean ± std) on three benchmark datasets under different settings.
See this image and copyright information in PMC

References

    1. Del Giorno A., Bagnell J.A., Hebert M. A discriminative framework for anomaly detection in large videos; Proceedings of the European Conference on Computer Vision; Amsterdam, The Netherlands. 8–16 October 2016; pp. 334–349.
    1. Tudor Ionescu R., Smeureanu S., Alexe B., Popescu M. Unmasking the abnormal events in video; Proceedings of the IEEE International Conference on Computer Vision; Venice, Italy. 22–29 October 2017; pp. 2895–2903.
    1. Wang S., Zeng Y., Liu Q., Zhu C., Zhu E., Yin J. Detecting abnormality without knowing normality: A two-stage approach for unsupervised video abnormal event detection; Proceedings of the 2018 ACM Multimedia Conference on Multimedia Conference (MM); Seoul, Korea. 22–26 October 2018; pp. 636–644.
    1. Chung P.C., Liu C.D. A daily behavior enabled hidden Markov model for human behavior understanding. Pattern Recognit. 2008;41:1572–1580. doi: 10.1016/j.patcog.2007.10.022. - DOI
    1. Foroughi H., Rezvanian A., Paziraee A. Robust fall detection using human shape and multi-class support vector machine; Proceedings of the Sixth Indian Conference on Computer Vision, Graphics & Image Processing (ICVGIP); Bhubaneswar, India. 16–19 December 2008; pp. 413–420.