A Content-Adaptively Sparse Reconstruction Method for Abnormal Events Detection With Low-Rank Property

Yu, Bosi; Liu, Yazhou; Sun, Quansen

doi:10.1109/tsmc.2016.2638048

Cited by 28 publications

(22 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ROC curves of the baseline methods are taken from the original papers (when available). From the frame-level evaluation results, it shows that our method outperforms most previous methods and yields competitive performance comparing with two state-of-the-art methods AMDN [21] and Sparse reconstruction [37]. Moreover, from the pixel-level evaluation results, which reflect the accuracy of anomaly localization, our method outperforms all the competing approaches.…”

Section: Qualitative and Quantitative Resultsmentioning

confidence: 81%

“…Specifically, we consider some classical methods that are widely cited as the baselines for the UCSD Anomaly Detection Dataset, which contains the sparse combination learning framework (SCLF) in [36], the mixture of probabilistic principal component analyzers (MPPCA) approach in [31], the social force model (SF) in [44], and their extension (SF+MPPCA) in [32], mixture of dynamic texture (MDT) in [32] and Adam method in [30]. In addition to these classical baselines, we also consider two state-of-the-art methods, the sparse reconstruction method in [37] and the Appearance and Motion DeepNet (AMDN) method in [21].…”

Section: Qualitative and Quantitative Resultsmentioning

confidence: 99%

“…In order to accelerate both the training and testing process, Lu et al [36] learned multiple dictionaries to encode normal size-invariant patches from multiscale frames. Yu et al [37] captured the low-rank property of the bases in dictionary learning phase, then a weighted sparse reconstruction method was used to measure the abnormality of testing samples.…”

Section: Hand-crafted Features Based Methodsmentioning

confidence: 99%

See 2 more Smart Citations

An Efficient Anomaly Detection System for Crowded Scenes Using Variational Autoencoders

Mei

Chen

et al. 2019

Applied Sciences

View full text Add to dashboard Cite

Anomaly detection in crowded scenes is an important and challenging part of the intelligent video surveillance system. As the deep neural networks make success in feature representation, the features extracted by a deep neural network represent the appearance and motion patterns in different scenes more specifically, comparing with the hand-crafted features typically used in the traditional anomaly detection approaches. In this paper, we propose a new baseline framework of anomaly detection for complex surveillance scenes based on a variational auto-encoder with convolution kernels to learn feature representations. Firstly, the raw frames series are provided as input to our variational auto-encoder without any preprocessing to learn the appearance and motion features of the receptive fields. Then, multiple Gaussian models are used to predict the anomaly scores of the corresponding receptive fields. Our proposed two-stage anomaly detection system is evaluated on the video surveillance dataset for a large scene, UCSD pedestrian datasets, and yields competitive performance compared with state-of-the-art methods.

show abstract

Section: Qualitative and Quantitative Resultsmentioning

confidence: 81%

Section: Qualitative and Quantitative Resultsmentioning

confidence: 99%

Section: Hand-crafted Features Based Methodsmentioning

confidence: 99%

See 1 more Smart Citation

An Efficient Anomaly Detection System for Crowded Scenes Using Variational Autoencoders

Mei

Chen

et al. 2019

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Levy et al [35] have developed a dataset that is meant for experimenting various identification of abnormal events in real-time. The work carried out by Yu et al [36] has used sparsity-based approach along with gradient feature to offer better reconstruction method. The next section discusses problems being identified from existing literature.…”

Section: Related Techniquesmentioning

confidence: 99%

A Novel Unsupervised Abnormal Event Identification Mechanism for Analysis of Crowded Scene

Pushpa¹

2017

ijacsa

View full text Add to dashboard Cite

Abstract-The advancement of visual sensing has introduced better capturing of the discrete information from a complex, crowded scene for assisting in the analysis. However, after reviewing existing system, we find that majority of the work carried out till date is associated with significant problems in modeling event detection as well as reviewing abnormality of the given scene. Therefore, the proposed system introduces a model that is capable of identifying the degree of abnormality for an event captured on the crowded scene using unsupervised training methodology. The proposed system contributes to developing a novel region-wise repository to extract the contextual information about the discrete-event for a given scene. The study outcome shows highly improved the balance between the computational time and overall accuracy as compared to the majority of the standard research work emphasizing on event detection.

show abstract

“…The OCC has been generally employed to solve the problem of novelty, outlier, intrusion, anomaly or fault detection [2]. These detection techniques are applicable for various types of applications across many disciplines [3]- [10]. Such types of problems can also be solved by multi-class classification when samples of both the classes, the normal and the outlier class, are available [11]- [15].…”

Section: Introductionmentioning

confidence: 99%

Adaptive Online Learning With Regularized Kernel for One-Class Classification

Gautam

Tiwari

Sundaram

et al. 2020

IEEE Trans. Syst. Man Cybern, Syst.

View full text Add to dashboard Cite

This paper presents adaptive Online learning with Regularized Kernel-based One-Class Extreme Learning Machine (ELM) classifiers for detection of outliers, and are collectively referred to as "ORK-OCELM". Two frameworks viz., boundary and reconstruction, are presented to detect the target class in ORK-OCELM. The kernel hyperplane based baseline one-class ELM model considers whole data in a single chunk, however, the proposed one-class classifiers are adapted in an online fashion from the stream of training samples.The performance of ORK-OCELM is evaluated on standard benchmark as well as synthetic datasets for both types of environment, i.e. stationary and non-stationary. While evaluating on stationary datasets, these classifiers are compared against batch learning based one-class classifiers. Similarly, while evaluating on non-stationary datasets, comparison is done with incremental learning based online one-class classifiers. The results indicate that the proposed classifiers yield better or similar outcomes for both. In the non-stationary dataset evaluation, adaptability of proposed classifiers in a changing environment is also demonstrated. It is further shown that proposed classifiers have large stream data handling capability even under limited system memory.Moreover, the proposed classifiers gain significant time improvement compared to traditional online one-class classifiers (in all aspects of training and testing). A faster learning ability of the proposed classifiers makes them more suitable for real-time anomaly detection.

show abstract

A Content-Adaptively Sparse Reconstruction Method for Abnormal Events Detection With Low-Rank Property

Cited by 28 publications

References 69 publications

An Efficient Anomaly Detection System for Crowded Scenes Using Variational Autoencoders

An Efficient Anomaly Detection System for Crowded Scenes Using Variational Autoencoders

A Novel Unsupervised Abnormal Event Identification Mechanism for Analysis of Crowded Scene

Adaptive Online Learning With Regularized Kernel for One-Class Classification

Contact Info

Product

Resources

About