Deep one-class classification variants for anomaly detection learn a mapping that concentrates nominal samples in feature space causing anomalies to be mapped away. Because this transformation is highly non-linear, finding interpretations poses a significant challenge. In this paper we present an explainable deep one-class classification method, Fully Convolutional Data Description (FCDD), where the mapped samples are themselves also an explanation heatmap. FCDD yields competitive detection performance and provides reasonable explanations on common anomaly detection benchmarks with CIFAR-10 and ImageNet. On MVTec-AD, a recent manufacturing dataset offering ground-truth anomaly maps, FCDD sets a new state of the art in the unsupervised setting. Our method can incorporate ground-truth anomaly maps during training and using even a few of these (~5) improves performance significantly. Finally, using FCDD's explanations we demonstrate the vulnerability of deep one-class classification models to spurious image features such as image watermarks.
In this paper, we propose a novel method for irregularity detection. Previous researches solve this problem as a One-Class Classification (OCC) task where they train a reference model on all of the available samples. Then, they consider a test sample as an anomaly if it has a diversion from the reference model. Generative Adversarial Networks (GANs) have achieved the most promising results for OCC while implementing and training such networks, especially for the OCC task, is a cumbersome and computationally expensive procedure. To cope with the mentioned challenges, we present a simple but effective method to solve the irregularity detection as a binary classification task in order to make the implementation easier along with improving the detection performance. We learn two deep neural networks (generator and discriminator) in a GAN-style setting on merely the normal samples. During training, the generator gradually becomes an expert to generate samples which are similar to the normal ones. In the training phase, when the generator fails to produce normal data (in the early stages of learning and also prior to the complete convergence), it can be considered as an irregularity generator. In this way, we simultaneously generate the irregular samples. Afterward, we train a binary classifier on the generated anomalous samples along with the normal instances in order to be capable of detecting irregularities. The proposed framework applies to different related applications of outlier and anomaly detection in images and videos, respectively. The results confirm that our proposed method is superior to the baseline and state-of-the-art solutions.
The traditional approach to face anti-spoofing sees it as a binary classification problem, and binary classifiers are trained and validated on specialized anti-spoofing databases. One of the drawbacks of this approach is that, due to the variability of face spoofing attacks, environmental factors, and the typically small sample size, such classifiers do not generalize well to previously unseen databases. Anomaly detection, which approaches face anti-spoofing as a one-class classification problem, is emerging as an increasingly popular alternative approach. Nevertheless, in all existing work on anomaly detection for face anti-spoofing, the proposed training protocols utilize images from specialized anti-spoofing databases only, even though only common images of real faces are needed. Here, we explore the use of in-the-wild images, and images from non-specialized face databases, to train one-class classifiers for face anti-spoofing. Employing a well-established technique, we train a convolutional autoencoder on real faces and compare the reconstruction error of the input against a threshold to classify a face image accordingly as either client or imposter. Our results show that the inclusion in the training set of in-the-wild images increases the discriminating power of the classifier significantly on an unseen database, as evidenced by a large increase in the value of the Area Under the Curve. In a limitation of our approach, we note that the problem of finding a suitable operating point on the unseen database remains a challenge, as evidenced by the values of the Half Total Error Rate.
Fraud detection is an important topic that applies to various enterprises such as banking and financial sectors, insurance, government agencies, law enforcement, and more. Fraud attempts have been risen remarkably in current years, shaping fraud detection an essential topic for research. One of the main challenges in fraud detection is acquiring fraud samples which is a complex and challenging task. In order to deal with this challenge, we apply one-class classification methods such as SVDD which does not need the fraud samples for training. Also, we present our algorithm REDBSCAN which is an extension of DBSCAN to reduce the number of samples and select those that keep the shape of data. The results obtained by the implementation of the proposed method indicated that the fraud detection process was improved in both performance and speed.
Machine learning systems and also, specifically, automatic speech recognition (ASR) systems are vulnerable against adversarial attacks, where an attacker maliciously changes the input. In the case of ASR systems, the most interesting cases are targeted attacks, in which an attacker aims to force the system into recognizing given target transcriptions in an arbitrary audio sample. The increasing number of sophisticated, quasi imperceptible attacks raises the question of countermeasures. In this paper, we focus on hybrid ASR systems and compare four acoustic models regarding their ability to indicate uncertainty under attack: a feed-forward neural network and three neural networks specifically designed for uncertainty quantification, namely a Bayesian neural network, Monte Carlo dropout, and a deep ensemble. We employ uncertainty measures of the acoustic model to construct a simple one-class classification model for assessing whether inputs are benign or adversarial. Based on this approach, we are able to detect adversarial examples with an area under the receiving operator curve score of more than 0.99. The neural networks for uncertainty quantification simultaneously diminish the vulnerability to the attack, which is reflected in a lower recognition accuracy of the malicious target text in comparison to a standard hybrid ASR system.
One critical prerequisite for the deployment of reinforcement learning systems in the real world is the ability to reliably detect situations on which the agent was not trained. Such situations could lead to potential safety risks when wrong predictions lead to the execution of harmful actions. In this work, we propose PEOC, a new policy entropy based out-of-distribution classifier that reliably detects unencountered states in deep reinforcement learning. It is based on using the entropy of an agent's policy as the classification score of a one-class classifier. We evaluate our approach using a procedural environment generator. Results show that PEOC is highly competitive against state-of-the-art one-class classification algorithms on the evaluated environments. Furthermore, we present a structured process for benchmarking out-of-distribution classification in reinforcement learning.
Dictionary learning for sparse representations has been successful in many reconstruction tasks. Simplicial learning is an adaptation of dictionary learning, where subspaces become clipped and acquire arbitrary offsets, taking the form of simplices. Such adaptation is achieved through additional constraints on sparse codes. Furthermore, an evolutionary approach can be chosen to determine the number and the dimensionality of simplices composing the simplicial, in which most generative and compact simplicials are favored. This paper proposes an evolutionary simplicial learning method as a generative and compact sparse framework for classification. The proposed approach is first applied on a one-class classification task and it appears as the most reliable method within the considered benchmark. Most surprising results are observed when evolutionary simplicial learning is considered within a multi-class classification task. Since sparse representations are generative in nature, they bear a fundamental problem of not being capable of distinguishing two classes lying on the same subspace. This claim is validated through synthetic experiments and superiority of simplicial learning even as a generative-only approach is demonstrated. Simplicial learning loses its superiority over discriminative methods in high-dimensional cases but can further be modified with discriminative elements to achieve state-of-the-art performance in classification tasks.
Inexpensive sensing and computation, as well as insurance innovations, have made smart dashboard cameras ubiquitous. Increasingly, simple model-driven computer vision algorithms focused on lane departures or safe following distances are finding their way into these devices. Unfortunately, the long-tailed distribution of road hazards means that these hand-crafted pipelines are inadequate for driver safety systems. We propose to apply data-driven anomaly detection ideas from deep learning to dashcam videos, which hold the promise of bridging this gap. Unfortunately, there exists almost no literature applying anomaly understanding to moving cameras, and correspondingly there is also a lack of relevant datasets. To counter this issue, we present a large and diverse dataset of truck dashcam videos, namely RetroTrucks, that includes normal and anomalous driving scenes. We apply: (i) one-class classification loss and (ii) reconstruction-based loss, for anomaly detection on RetroTrucks as well as on existing static-camera datasets. We introduce formulations for modeling object interactions in this context as priors. Our experiments indicate that our dataset is indeed more challenging than standard anomaly detection datasets, and previous anomaly detection methods do not perform well here out-of-the-box. In addition, we share insights into the behavior of these two important families of anomaly detection approaches on dashcam data.
Cluster of viral pneumonia occurrences during a short period of time may be a harbinger of an outbreak or pandemic, like SARS, MERS, and recent COVID-19. Rapid and accurate detection of viral pneumonia using chest X-ray can be significantly useful in large-scale screening and epidemic prevention, particularly when other chest imaging modalities are less available. Viral pneumonia often have diverse causes and exhibit notably different visual appearances on X-ray images. The evolution of viruses and the emergence of novel mutated viruses further result in substantial dataset shift, which greatly limits the performance of classification approaches. In this paper, we formulate the task of differentiating viral pneumonia from non-viral pneumonia and healthy controls into an one-class classification-based anomaly detection problem, and thus propose the confidence-aware anomaly detection (CAAD) model, which consists of a shared feature extractor, an anomaly detection module, and a confidence prediction module. If the anomaly score produced by the anomaly detection module is large enough or the confidence score estimated by the confidence prediction module is small enough, we accept the input as an anomaly case (i.e., viral pneumonia). The major advantage of our approach over binary classification is that we avoid modeling individual viral pneumonia classes explicitly and treat all known viral pneumonia cases as anomalies to reinforce the one-class model. The proposed model outperforms binary classification models on the clinical X-VIRAL dataset that contains 5,977 viral pneumonia (no COVID-19) cases, 18,619 non-viral pneumonia cases, and 18,774 healthy controls.
This paper presents a new approach to classification of high dimensional spectroscopy data and demonstrates that it outperforms other current state-of-the art approaches. The specific task we consider is identifying whether samples contain chlorinated solvents or not, based on their Raman spectra. We also examine robustness to classification of outlier samples that are not represented in the training set (negative outliers). A novel application of a locally-connected neural network (NN) for the binary classification of spectroscopy data is proposed and demonstrated to yield improved accuracy over traditionally popular algorithms. Additionally, we present the ability to further increase the accuracy of the locally-connected NN algorithm through the use of synthetic training spectra and we investigate the use of autoencoder based one-class classifiers and outlier detectors. Finally, a two-step classification process is presented as an alternative to the binary and one-class classification paradigms. This process combines the locally-connected NN classifier, the use of synthetic training data, and an autoencoder based outlier detector to produce a model which is shown to both produce high classification accuracy, and be robust to the presence of negative outliers.
In this paper, we propose a novel method for transforming data into a low-dimensional space optimized for one-class classification. The proposed method iteratively transforms data into a new subspace optimized for ellipsoidal encapsulation of target class data. We provide both linear and non-linear formulations for the proposed method. The method takes into account the covariance of the data in the subspace; hence, it yields a more generalized solution as compared to Subspace Support Vector Data Description for a hypersphere. We propose different regularization terms expressing the class variance in the projected space. We compare the results with classic and recently proposed one-class classification methods and achieve better results in the majority of cases. The proposed method is also noticed to converge much faster than recently proposed Subspace Support Vector Data Description.
To train a robust deep learning model, one usually needs a balanced set of categories in the training data. The data acquired in a medical domain, however, frequently contains an abundance of healthy patients, versus a small variety of positive, abnormal cases. Moreover, the annotation of a positive sample requires time consuming input from medical domain experts. This scenario would suggest a promise for one-class classification type approaches. In this work we propose a general one-class classification model for histology, that is meta-trained on multiple histology datasets simultaneously, and can be applied to new tasks without expensive re-training. This model could be easily used by pathology domain experts, and potentially be used for screening purposes.
Classical approaches for one-class problems such as one-class SVM and isolation forest require careful feature engineering when applied to structured domains like images. State-of-the-art methods aim to leverage deep learning to learn appropriate features via two main approaches. The first approach based on predicting transformations (Golan & El-Yaniv, 2018; Hendrycks et al., 2019a) while successful in some domains, crucially depends on an appropriate domain-specific set of transformations that are hard to obtain in general. The second approach of minimizing a classical one-class loss on the learned final layer representations, e.g., DeepSVDD (Ruff et al., 2018) suffers from the fundamental drawback of representation collapse. In this work, we propose Deep Robust One-Class Classification (DROCC) that is both applicable to most standard domains without requiring any side-information and robust to representation collapse. DROCC is based on the assumption that the points from the class of interest lie on a well-sampled, locally linear low dimensional manifold. Empirical evaluation demonstrates that DROCC is highly effective in two different one-class problem settings and on a range of real-world datasets across different domains: tabular data, images (CIFAR and ImageNet), audio, and time-series, offering up to 20% increase in accuracy over the state-of-the-art in anomaly detection. Code is available at https://github.com/microsoft/EdgeML.
Insect monitoring is crucial for understanding the consequences of rapid ecological changes, but taxa identification currently requires tedious manual expert work and cannot be scaled-up efficiently. Deep convolutional neural networks (CNNs), provide a viable way to significantly increase the biomonitoring volumes. However, taxa abundances are typically very imbalanced and the amounts of training images for the rarest classes are simply too low for deep CNNs. As a result, the samples from the rare classes are often completely missed, while detecting them has biological importance. In this paper, we propose combining the trained deep CNN with one-class classifiers to improve the rare species identification. One-class classification models are traditionally trained with much fewer samples and they can provide a mechanism to indicate samples potentially belonging to the rare classes for human inspection. Our experiments confirm that the proposed approach may indeed support moving towards partial automation of the taxa identification task.
Nowadays, graph-structured data are increasingly used to model complex systems. Meanwhile, detecting anomalies from graph has become a vital research problem of pressing societal concerns. Anomaly detection is an unsupervised learning task of identifying rare data that differ from the majority. As one of the dominant anomaly detection algorithms, One Class Support Vector Machine has been widely used to detect outliers. However, those traditional anomaly detection methods lost their effectiveness in graph data. Since traditional anomaly detection methods are stable, robust and easy to use, it is vitally important to generalize them to graph data. In this work, we propose One Class Graph Neural Network (OCGNN), a one-class classification framework for graph anomaly detection. OCGNN is designed to combine the powerful representation ability of Graph Neural Networks along with the classical one-class objective. Compared with other baselines, OCGNN achieves significant improvements in extensive experiments.
Anomaly detection algorithms find extensive use in various fields. This area of research has recently made great advances thanks to deep learning. A recent method, the deep Support Vector Data Description (deep SVDD), which is inspired by the classic kernel-based Support Vector Data Description (SVDD), is capable of simultaneously learning a feature representation of the data and a data-enclosing hypersphere. The method has shown promising results in both unsupervised and semi-supervised settings. However, deep SVDD suffers from hypersphere collapse -- also known as mode collapse, if the architecture of the model does not comply with certain architectural constraints, e.g. the removal of bias terms. These constraints limit the adaptability of the model and in some cases, may affect the model performance due to learning sub-optimal features. In this work, we consider two regularizers to prevent hypersphere collapse in deep SVDD. The first regularizer is based on injecting random noise via the standard cross-entropy loss. The second regularizer penalizes the minibatch variance when it becomes too small. Moreover, we introduce an adaptive weighting scheme to control the amount of penalization between the SVDD loss and the respective regularizer. Our proposed regularized variants of deep SVDD show encouraging results and outperform a prominent state-of-the-art method on a setup where the anomalies have no apparent geometrical structure.
In this paper we consider authentication at the physical layer, in which the authenticator aims at distinguishing a legitimate supplicant from an attacker on the basis of the characteristics of a set of parallel wireless channels, which are affected by time-varying fading. Moreover, the attacker's channel has a spatial correlation with the supplicant's one. In this setting, we assess and compare the performance achieved by different approaches under different channel conditions. We first consider the use of two different statistical decision methods, and we prove that using a large number of references (in the form of channel estimates) affected by different levels of time-varying fading is not beneficial from a security point of view. We then consider classification methods based on machine learning. In order to face the worst case scenario of an authenticator provided with no forged messages during training, we consider one-class classifiers. When instead the training set includes some forged messages, we resort to more conventional binary classifiers, considering the cases in which such messages are either labelled or not. For the latter case, we exploit clustering algorithms to label the training set. The performance of both nearest neighbor (NN) and support vector machine (SVM) classification techniques is evaluated. Through numerical examples, we show that under the same probability of false alarm, one-class classification (OCC) algorithms achieve the lowest probability of missed detection when a small spatial correlation exists between the main channel and the adversary one, while statistical methods are advantageous when the spatial correlation between the two channels is large.
This paper describes a new form of unsupervised learning, whose input is a set of unlabeled points that are assumed to be local maxima of an unknown value function v in an unknown subset of the vector space. Two functions are learned: (i) a set indicator c, which is a binary classifier, and (ii) a comparator function h that given two nearby samples, predicts which sample has the higher value of the unknown function v. Loss terms are used to ensure that all training samples x are a local maxima of v, according to h and satisfy c(x)=1. Therefore, c and h provide training signals to each other: a point x' in the vicinity of x satisfies c(x)=-1 or is deemed by h to be lower in value than x. We present an algorithm, show an example where it is more efficient to use local maxima as an indicator function than to employ conventional classification, and derive a suitable generalization bound. Our experiments show that the method is able to outperform one-class classification algorithms in the task of anomaly detection and also provide an additional signal that is extracted in a completely unsupervised way.
Robustness to out-of-distribution (OOD) data is an important goal in building reliable machine learning systems. Especially in autonomous systems, wrong predictions for OOD inputs can cause safety critical situations. As a first step towards a solution, we consider the problem of detecting such data in a value-based deep reinforcement learning (RL) setting. Modelling this problem as a one-class classification problem, we propose a framework for uncertainty-based OOD classification: UBOOD. It is based on the effect that an agent's epistemic uncertainty is reduced for situations encountered during training (in-distribution), and thus lower than for unencountered (OOD) situations. Being agnostic towards the approach used for estimating epistemic uncertainty, combinations with different uncertainty estimation methods, e.g. approximate Bayesian inference methods or ensembling techniques are possible. We further present a first viable solution for calculating a dynamic classification threshold, based on the uncertainty distribution of the training data. Evaluation shows that the framework produces reliable classification results when combined with ensemble-based estimators, while the combination with concrete dropout-based estimators fails to reliably detect OOD situations. In summary, UBOOD presents a viable approach for OOD classification in deep RL settings by leveraging the epistemic uncertainty of the agent's value function.
Data-driven fault diagnostics of safety-critical systems often faces the challenge of a complete lack of labeled data associated with faulty system conditions (i.e., fault types) at training time. Since an unknown number and nature of fault types can arise during deployment, data-driven fault diagnostics in this scenario is an open-set learning problem. Most of the algorithms for open-set diagnostics are one-class classification and unsupervised algorithms that do not leverage all the available labeled and unlabeled data in the learning algorithm. As a result, their fault detection and segmentation performance (i.e., identifying and separating faults of different types) are sub-optimal. With this work, we propose training a variational autoencoder (VAE) with labeled and unlabeled samples while inducing implicit supervision on the latent representation of the healthy conditions. This, together with a modified sampling process of VAE, creates a compact and informative latent representation that allows good detection and segmentation of unseen fault types using existing one-class and clustering algorithms. We refer to the proposed methodology as "knowledge induced variational autoencoder with adaptive sampling" (KIL-AdaVAE). The fault detection and segmentation capabilities of the proposed methodology are demonstrated in a new simulated case study using the Advanced Geared Turbofan 30000 (AGTF30) dynamical model under real flight conditions. In an extensive comparison, we demonstrate that the proposed method outperforms other learning strategies (supervised learning, supervised learning with embedding and semi-supervised learning) and deep learning algorithms, yielding significant performance improvements on fault detection and fault segmentation.