Anomaly detection approach using Hidden Markov Model

Dorj, Enkhjargal; Altangerel, Erdenebaatar

doi:10.1109/ifost.2013.6616874

Cited by 12 publications

(4 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For statistical-based approaches (i.e. methods based on statistical models postulated over the normal context, where the abnormal events are classified based on probabilities), there are, for instance: Gaussian mixture model [17] and hidden Markov model [18]. For pattern representation based on spatio-temporal, there are techniques such as histogram of oriented gradients [19], histogram of optical flow [20], textures of optical flow [21], tracking-based [22], and spatio-temporal texture [23].…”

Section: B Pattern Representationmentioning

confidence: 99%

Learning Spatio-Temporal Features for Detecting Anomalies in Videos using Convolutional Autoencoder

Ribeiro¹,

Romero²,

Lazzaretti³

et al. 2020

Anais Do 14. Congresso Brasileiro De Inteligência Computacional

View full text Add to dashboard Cite

Automatic video surveillance systems are a recurrent topic in recent video analysis research. Anomaly detection is an interesting way for tackling this problem, because video analysis is tedious and exhaustive for humans. Depending on the application field, anomalies can present different characteristics and challenges for pattern representation, requiring the design of hand-crafted features (such as spatial and temporal information). Deep learning methods have achieved the state-of-the-art performance for many recognition problems in recent years and may be an interesting choice for learning features automatically, since it captures the 2D structure in image sequences during the learning process. The deep Convolutional Autoencoder (CAE) may be an interesting approach for anomaly detection since they can learn signatures automatically in an unsupervised way. This work purposes the use of a deep CAE in the anomaly detection context for learning spatio-temporal signatures from raw video frames. Similar our previous work, we use as anomaly score a successful strategy based on the reconstruction error of a package of frames. The proposed methods were evaluated by means of several experiments with public-domain datasets. The promising results support further research in this area.

show abstract

Section: B Pattern Representationmentioning

confidence: 99%

Learning Spatio-Temporal Features for Detecting Anomalies in Videos using Convolutional Autoencoder

Ribeiro¹,

Romero²,

Lazzaretti³

et al. 2020

Anais Do 14. Congresso Brasileiro De Inteligência Computacional

View full text Add to dashboard Cite

show abstract

“…In response to this, unsupervised approaches have been widely explored and in the DCASE 2020 task 2 challenge they have showcased promising results [9,32]. Researchers used the Hidden Markov Model (HMM) [33], Gaussian Mixture Model (GMM) [34], Nonnegative Matrix Factorization (NMF) [35], and deep learning-based approaches, such as autoencoders and Generative models [36], to model and analyze normal sounds. By training these models to compress and reconstruct normal sounds, the underlying properties of normal sounds were learned effectively in the latent space, and when an abnormal sample was introduced into these models, the resulting large reconstruction errors indicated its deviation from the learned normal sound distribution.…”

Section: Introductionmentioning

confidence: 99%

Regularized Contrastive Masked Autoencoder Model for Machinery Anomaly Detection Using Diffusion-Based Data Augmentation

Zahedi,

Saraee,

Masoumi

et al. 2023

Algorithms

View full text Add to dashboard Cite

Unsupervised anomalous sound detection, especially self-supervised methods, plays a crucial role in differentiating unknown abnormal sounds of machines from normal sounds. Self-supervised learning can be divided into two main categories: Generative and Contrastive methods. While Generative methods mainly focus on reconstructing data, Contrastive learning methods refine data representations by leveraging the contrast between each sample and its augmented version. However, existing Contrastive learning methods for anomalous sound detection often have two main problems. The first one is that they mostly rely on simple augmentation techniques, such as time or frequency masking, which may introduce biases due to the limited diversity of real-world sounds and noises encountered in practical scenarios (e.g., factory noises combined with machine sounds). The second issue is dimension collapsing, which leads to learning a feature space with limited representation. To address the first shortcoming, we suggest a diffusion-based data augmentation method that employs ChatGPT and AudioLDM. Also, to address the second concern, we put forward a two-stage self-supervised model. In the first stage, we introduce a novel approach that combines Contrastive learning and masked autoencoders to pre-train on the MIMII and ToyADMOS2 datasets. This combination allows our model to capture both global and local features, leading to a more comprehensive representation of the data. In the second stage, we refine the audio representations for each machine ID by employing supervised Contrastive learning to fine-tune the pre-trained model. This process enhances the relationship between audio features originating from the same machine ID. Experiments show that our method outperforms most of the state-of-the-art self-supervised learning methods. Our suggested model achieves an average AUC and pAUC of 94.39% and 87.93% on the DCASE 2020 Challenge Task2 dataset, respectively.

show abstract

“…In the study conducted by Dorj and Altangerel (Dorj and Altangerel 2013), using the Discrete Hidden Markov Model, anomaly detection is made for discrete sequences. As a result of the tests conducted it is seen that the method works on the discrete data with 85.7% of correctness.…”

Section: Introductionmentioning

confidence: 99%

“…There are many parameters that can be used to determine different density states of the network. These parameters can be listed basically as response time, connection time, throughput, jitter for video data, retransmission count and delay (Dorj and Altangerel 2013). For the analysis of the network structure given in the article the parameters, throughput, delay, retransmission count and response time are used.…”

mentioning

confidence: 99%

Untitled

Gülağız

Gök

Şahin

2018

KSEJ

View full text Add to dashboard Cite

abnormal circumstances while monitoring the systems and target to take the necessary precaution. Attack detection mechanisms used by IDS are basically divided into two. These are signature-based methods and anomaly based methods. Signature-based methods are used more for known intrusion types prevalently. Anomaly based methods model the normal behavior of the network and determines abnormal circumstances through the obtained model of normal state. Anomaly based methods are preferred more as they are more successful than the signature-based model in terms of detecting the new intrusion types. Anomaly detection systems determine behaviors that show deviation from the expected normal usage profiles as an anomaly. In the detection of the anomaly the normal behavior of the system is generally obtained using the statistical methods. One of the statistical methods used generally for this purpose is the Hidden Markov Model (HMM). Markov model is used to represent the tiered observations happen in time. HMM is a special version of the Markov

show abstract

Anomaly detection approach using Hidden Markov Model

Cited by 12 publications

References 6 publications

Learning Spatio-Temporal Features for Detecting Anomalies in Videos using Convolutional Autoencoder

Learning Spatio-Temporal Features for Detecting Anomalies in Videos using Convolutional Autoencoder

Regularized Contrastive Masked Autoencoder Model for Machinery Anomaly Detection Using Diffusion-Based Data Augmentation

Untitled

Contact Info

Product

Resources

About