Subspace Support Vector Data Description

Sohrab, Fahad; Raitoharju, Jenni; Gabbouj, Moncef; Iosifidis, Alexandros

doi:10.1109/icpr.2018.8545819

Cited by 48 publications

(26 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It thus explains why anomaly detection can benefit from minimizing the reconstruction error. Link to one-class classification based approaches: Some one-class classification based methods [4,14,15] minimize the volume of a data-enclosing hyper-sphere in latent space, which is mathematically shown equivalent to minimizing the upper bound on the entropy of latent space [16]. Thus this kind of method can be linked to optimizing the entropy term H n (z) in Eq.…”

Section: Relation To Existing Algorithmsmentioning

confidence: 99%

Deep Unsupervised Image Anomaly Detection: An Information Theoretic Framework

Zheng

Huang

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

Surrogate task based methods have recently shown great promise for unsupervised image anomaly detection. However, there is no guarantee that the surrogate tasks share the consistent optimization direction with anomaly detection. In this paper, we return to a direct objective function for anomaly detection with information theory, which maximizes the distance between normal and anomalous data in terms of the joint distribution of images and their representation. To make this objective function directly optimizable under the unsupervised setting, we manage to find its lower bound which weights the trade-off between mutual information and entropy, which leads to a novel information theoretic framework for unsupervised image anomaly detection. Extensive experiments on several benchmark data sets have shown that the proposed framework significantly outperforms several state-of-the-arts.

show abstract

Section: Relation To Existing Algorithmsmentioning

confidence: 99%

Deep Unsupervised Image Anomaly Detection: An Information Theoretic Framework

Zheng

Huang

et al. 2021

2021 IEEE International Conference on Image Processing (ICIP)

View full text Add to dashboard Cite

show abstract

“…Subsequent extensions of SVDD and OC-SVM were proposed to improve their accuracy such as the Graph Embedded OC-SVM (GE-OC-SVM) and graph embedded SVDD (GE-SVDD) 27 or, more recently, the subspace-SVDD (S-SVDD). 28 The first approach based on deep learning is the deep-SVDD. 29 Alternatively, there are other approaches that employ a set of ellipsoids to fit the region of the data space, 30,31 or others that employ a family of convex hulls for one-class classification 32 like the well-known approximate polytope ensemble algorithm (APE).…”

Section: Related Workmentioning

confidence: 99%

“…Later on, several more popular methods were developed like support vector data description (SVDD) 25 or the one‐class support vector machine (OC‐SVM), 26 that has become one of the most successful tool for one‐class classification. Subsequent extensions of SVDD and OC‐SVM were proposed to improve their accuracy such as the Graph Embedded OC‐SVM (GE‐OC‐SVM) and graph embedded SVDD (GE‐SVDD) 27 or, more recently, the subspace‐SVDD (S‐SVDD) 28 . The first approach based on deep learning is the deep‐SVDD 29 .…”

Section: Related Workmentioning

confidence: 99%

DSVD‐autoencoder: A scalable distributed privacy‐preserving method for one‐class classification

2020

View full text Add to dashboard Cite

One-class classification has gained interest as a solution to certain kinds of problems typical in a wide variety of real environments like anomaly or novelty detection. Autoencoder is the type of neural network that has been widely applied in these one-class problems. In the Big Data era, new challenges have arisen, mainly related with the data volume. Another main concern derives from Privacy issues when data is distributed and cannot be shared among locations. These two conditions make many of the classic and brilliant methods not applicable. In this paper, we present distributed singular value decomposition (DSVDautoencoder), a method for autoencoders that allows learning in distributed scenarios without sharing raw data. Additionally, to guarantee privacy, it is noniterative and hyperparameter-free, two interesting characteristics when dealing with Big Data. In comparison with the state of the art, results demonstrate that DSVD-autoencoder provides a highly competitive solution to deal with very large data sets by reducing training from several hours to seconds while maintaining good accuracy.

show abstract

“…where, as in OC-SVM, the ξ's model the slack. There have been extensions of this scheme, such as the mSVDD that uses a mixture of such hyperspheres [29], density-induced SVDD [30], using kernelized variants [52], and more recently, to use subspaces for data description [49]. A major drawback of SVDD in general is the strong assumption it makes on the isotropic nature of the underlying data distribution.…”

Section: Background and Related Workmentioning

confidence: 99%

GODS: Generalized One-Class Discriminative Subspaces for Anomaly Detection

Wang

Cherian

2019

2019 IEEE/CVF International Conference on Computer Vision (ICCV)

101

View full text Add to dashboard Cite

One-class learning is the classic problem of fitting a model to data for which annotations are available only for a single class. In this paper, we propose a novel objective for one-class learning. Our key idea is to use a pair of orthonormal frames -as subspaces -to "sandwich" the labeled data via optimizing for two objectives jointly: i) minimize the distance between the origins of the two subspaces, and ii) to maximize the margin between the hyperplanes and the data, either subspace demanding the data to be in its positive and negative orthant respectively. Our proposed objective however leads to a non-convex optimization problem, to which we resort to Riemannian optimization schemes and derive an efficient conjugate gradient scheme on the Stiefel manifold.To study the effectiveness of our scheme, we propose a new dataset Dash-Cam-Pose, consisting of clips with skeleton poses of humans seated in a car, the task being to classify the clips as normal or abnormal; the latter is when any human pose is out-of-position with regard to say an airbag deployment. Our experiments on the proposed Dash-Cam-Pose dataset, as well as several other standard anomaly/novelty detection benchmarks demonstrate the benefits of our scheme, achieving state-of-the-art oneclass accuracy. 5 The data follows the formula f (x) = √ x * (x + sign(randn) * rand), where randn and rand are standard MATLAB functions.

show abstract

Subspace Support Vector Data Description

Cited by 48 publications

References 78 publications

Deep Unsupervised Image Anomaly Detection: An Information Theoretic Framework

Deep Unsupervised Image Anomaly Detection: An Information Theoretic Framework

DSVD‐autoencoder: A scalable distributed privacy‐preserving method for one‐class classification

GODS: Generalized One-Class Discriminative Subspaces for Anomaly Detection

Contact Info

Product

Resources

About