Research On Unsupervised Video Anomaly Detection Algorithms And Applications

With an increasing number of videos collected from various surveillance cameras,intelligent video surveillance systems becomes more and more urgent in social society especially for public security service and traffic management.In modern intelligent video surveillance systems,anomaly detection plays an essential role which not only significantly increases monitoring efficiency especially in security area but also alleviates the working burden of traditional video surveillance system operators.Almost all existing methods tackle the problem under the supervised setting,which will cost a lot of labor costs,materials,time for data annotation.Fully unsupervised manner can overcome those shortcomings,but only a few attempts are explored without any prior information.This paper focuses on unsupervised video anomaly detection technology and application research,which can be summarized as following:(1)We bring up a method called two-stage unsupervised video anomaly detection using low-rank based unsupervised one-class learning with ridge regression.To avoid the cost of labeling training videos,we propose to discriminate anomaly by a novel two-stage framework in a fully unsupervised manner.Unlike previous unsupervised approaches using local change detection to discover abnormality,our method enjoys the global information from video context by considering the pairwise similarity of all video events.In this way,our algorithm formulates video anomaly detection as an extension of unsupervised one-class learning.Specifically,our method consists of two stages: The first stage of our kernel-based method,named LR-UOCL-RR,reformulates the optimization goal of UOCL with ridge regression to avoid expensive computation,which enables our method to handle massive unlabeled data from videos.In the second stage,the estimated normal video events from the first stage are fed into the one-class support vector machine to refine the profile around normal events and enhance the performance.The experimental results conducted on two challenging video benchmarks indicate that our method is superior,up to 15.7% AUC gain,to the state-of-the-art methods in the unsupervised anomaly detection task and even better than several supervised approaches.(2)We propose an efficient and robust unsupervised anomaly detection method using ensemble random projection in surveillance videos.To avoid massive computation caused by back-propagation in existing methods,we utilize an 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.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.