Interestingness Prediction by Robust Learning to Rank

Unsupervised outlier detection using random subspace and subsampling ensembles of Dirichlet process mixtures

Restricted default theories: Expressive power and outlier detection tasks

Person search over security video surveillance systems using deep learning methods: A review

Over the last few years, Kernel Principal Component Analysis (KPCA) has found several applications in outlier detection. A relatively recent method uses KPCA to compute the reconstruction error (RE) of previously unseen samples and, via thresholding, to identify atypical samples. In this paper we propose an alternative method, which performs the same task, but considers Mahalanobis distances in the orthogonal complement of the subspace that is utilized to compute the reconstruction error. In order to illustrate its merits, we provide qualitative and quantitative results on both artificial and real datasets and we show that it is competitive, if not superior, for several outlier detection tasks, when compared to the original RE-based variant and the One-Class SVM detection approach.

Converting SVDD scores into probability estimates: Application to outlier detection

In recent years, anomaly detection techniques in time-series data have been widely used in manufacturing, cybersecurity, and other fields. Meanwhile, various anomaly detection models based on generative adversarial networks (GAN) are gradually used in time-series anomaly detection tasks. However, there are problems of unstable generator training, missed detection of anomalous data, and inconsistency between the discriminator's discriminant and the anomaly detection target in GAN networks. Aiming at the above problems, the paper proposes a DUAL-ADGAN (Dual Anomaly Detection Generative Adversarial Networks) model for the detection of anomalous data in time series. First, the Wasserstein distance satisfying the Lipschitz constraint is used as the loss function of the data reconstruction module, which improves the stability of the traditional GAN network training. Second, by adding a data prediction module to the DUAL-ADGAN model, the distinction between abnormal and normal samples is increased, and the rate of missing abnormal data in the model is reduced. Third, by introducing the Fence-GAN loss function, the discriminator is aligned with the anomaly detection target, which effectively reduces the anomaly data false detection rate of the DUAL-ADGAN model. Finally, anomaly scores derived from the DUAL-ADGAN model are compared with dynamic thresholds to detect anomalies. The experimental results show that the average F1 of the DUAL-ADGAN model is 0.881, which is better than the other nine baseline models. The conclusions demonstrate that the DUAL-ADGAN model proposed in the paper is more stable in training while effectively solving the problems of anomaly miss detection and discriminator inconsistency with the anomaly detection target in the anomaly detection task.

Spatiotemporal data mining (STDM) discovers useful patterns from the dynamic interplay between space and time. Several available surveys capture STDM advances and report a wealth of important progress in this field. However, STDM challenges and problems are not thoroughly discussed and presented in articles of their own. We attempt to fill this gap by providing a comprehensive literature survey on state-of-the-art advances in STDM. We describe the challenging issues and their causes and open gaps of multiple STDM directions and aspects. Specifically, we investigate the challenging issues in regards to spatiotemporal relationships, interdisciplinarity, discretisation, and data characteristics. Moreover, we discuss the limitations in the literature and open research problems related to spatiotemporal data representations, modelling and visualisation, and comprehensiveness of approaches. We explain issues related to STDM tasks of classification, clustering, hotspot detection, association and pattern mining, outlier detection, visualisation, visual analytics, and computer vision tasks. We also highlight STDM issues related to multiple applications including crime and public safety, traffic and transportation, earth and environment monitoring, epidemiology, social media, and Internet of Things.

How to properly involve text characteristics like multi-scale, arbitrary direction, length aspect ratio, into detection network design has become a hot topic in computer vision. Feature Pyramid Network (FPN) is a typical method to achieve robust text detection, where its low-level and high-level feature map retains spatial structure and global semantic information, respectively. However, its strict hierarchical structure fails to fuse low-level and high-level information to improve the distinguish ability of feature map. To address this problem, we propose a novel feature fusion pyramid network for end-to-end scene text detection by fusing multi-modal information. By dividing pyramid structure into high-level and low-level layers, channel and spatial attention modules are adopted to enhance high-level and low-level feature representation by encoding channel and spatial-wise context information, respectively. In order to reduce information loss by layer transmission, a special residual network is designed to achieve short-cut between high-level and low-level features, so as to realize multi-modal feature fusion. Experiments show the precision and recall of the proposed method on ICDAR2015, ICDAR2017-MLT, and MSRA-TD500 datasets reach 88.7%/82.1%, 77.0%/60.3%, and 85.3%/74.8%, respectively.

Shaft orbit plays an important role in condition monitoring and fault diagnosis for hydropower unit. A novel method of shaft orbit identification based on low-level image feature representation and classification is proposed. The main characteristic is that the vibrations of the shaft in terms of displacements are used to draw points in an image panel at a fixed scale, resulting in the shaft orbit image set. Histogram of Oriented Gradients (HOG) is then used as the low-level local shape descriptor. Accordingly, a given shaft orbit image can be represented by a plenty of HOG local descriptors which are further aggregated into a feature vector. The feature vectors associated with class labels are fed to linear classifiers for multi-class classification. To deal with noisy samples robustly and solve the problem that training samples always cannot be separated in original space, kernel-based soft-margin Support Vector Machine (SVM) is employed. The proposed algorithm is implemented and tested on the challenging data set which is collected from a testing apparatus under different fault settings. It yields a satisfactory recognition rate which is 98.35% on the overall data set.

The limited availability of high-quality training data poses a persistent challenge for synthetic aperture radar (SAR) target classification. Existing data augmentation methods mainly adopt a simplistic application of GAN-based style transfer techniques to directly synthesize pseudo-SAR images from optical images. However, our in-depth analysis of this cross-modal conversion reveals that such straightforward strategies primarily focus on transferring high-level semantic information (e.g., target shapes), thus failing to adequately capture the essential low-level features unique to SAR imagery (e.g., scattering textures). To address this inherent trade-off between high-level semantic preservation and low-level feature authenticity, we propose a Hierarchical Feature-Constrained GAN (HiFC-GAN) tailored for optical-to-SAR style transfer. Specifically, HiFC-GAN enhances the representation of low-level SAR features by introducing local texture contrast constraints at shallow layers, while introducing explicit feature mapping constraints at deeper layers to maintain high-level semantic consistency throughout the reconstruction process. Experimental results demonstrate that HiFC-GAN significantly outperforms existing GAN-based techniques in image generation quality, particularly improving the low-level feature authenticity of pseudo-SAR images. Moreover, the generated pseudo-SAR images further improve the performance of downstream target classification tasks, yielding accuracy gains ranging from 3.56% to 5.90% on average with mainstream CNN-based models.

Anomaly detection aims to detect abnormal samples from normal data. Existing methods mainly use synthesis-based approach and uncertainty-based approach. Moreover, the above-mentioned methods need to re-train part/all of the deep network, which limits its application scenarios especially real-time scenarios. We propose a novel retraining-free anomaly detection method, which includes a dual-branch anomaly detection mechanism to detect abnormal pixels from two views. In the pseudo-supervised branch, we use the Copy post method to transform the unsupervised anomaly detection task into a supervised binary detection task. In the prototype detection branch, we use prototypes to improve the accuracy of abnormal pixel detection. The anomaly results of the two branches are fused by means of soft voting to produce anomaly detection results with higher recall. Our experiments show that our method significantly outperforms other methods and achieves excellent results on different datasets, which also shows that our method has good generalization ability. Therefore, our proposed method can effectively improve the effectiveness and safety of anomaly detection in real-time scenarios, such as autonomous driving.

Anomaly detection is often approached as an out-of-distribution (OOD) detection task, where a feature distribution from normal samples is constructed, and deviations are flagged as anomalies. This approach is dependent on manual labeling, as subtle visual anomalies can be easily overlooked, resulting in the potential for bias in labeling and subsequent unsatisfactory detection results. Based on the issue, we propose Anomaly Detection with Succinct Feature Filtering (ADSFF) for unlabeled samples. Our method avoids sample labeling bias and provides a solution to the coexistence of anomalous and normal features in the feature space of unlabeled samples. ADSFF includes a data preprocessing module and a feature filtering module, where the data preprocessing module improves the visibility of subtle anomalies, while the feature filtering module screens the local features of the samples. In feature filtering, we found that feedforward neural networks do not lose feature information during the feature transformation process. Consequently, we utilized feedforward neural networks for feature filtering and achieved expected results. Furthermore, we investigate the impact of sample imbalance on the task of anomaly detection using unlabeled samples. This paper assesses the performance of ADSFF using the MVTec AD and BeanTech Anomaly Detection (BTAD) datasets. The results demonstrate that ADSFF achieves an average area under the curve (AUC) of 0.978 on the MVTec AD and an average AUC of 0.942 on the BTAD. ADSFF outperformed other methods on seven test datasets in MVTec AD, achieving the highest average accuracy on MVTec AD.

Anomaly detection is a classic and crucial problem in the field of artificial intelligence, which aims to find instances that deviate so much from the main distribution of data or differ from known instances. This paper explores how to combine advanced deep learning techniques with traditional probabilistic statistical methods for anomaly detection. We propose a very effective and concise semi-supervised anomaly detection method named “ORGAN-KDE” based on the orthogonal generative adversarial network (O-GAN) and kernel density estimation. In the training phase, we use the encoder of O-GAN to learn the latent representation of normal data, namely the code of normal data, and then use the kernel density estimation to solve the probability density function of code. The code of normal sample obtained through the trained encoder can get a larger probability value when passing through the trained kernel density estimator, while the code of anomalous sample can get a smaller probability value, so as to achieve the purpose of anomaly detection. Compared with other anomaly detection methods based on GAN, our method has a very simple network structure, and experiments have proved that it performs well on both structured datasets and image datasets.

Supervised deep learning approaches provide state-of-the-art performance on medical image classification tasks for disease screening. However, these methods require large labeled datasets that involve resource-intensive expert annotation. Further, disease screening applications have low prevalence of abnormal samples; this class imbalance makes the task more akin to anomaly detection. While the machine learning community has proposed unsupervised deep learning methods for anomaly detection, they have yet to be characterized on medical images where normal vs. anomaly distinctions may be more subtle and variable. In this work, we characterize existing unsupervised anomaly detection methods on retinal fundus images, and find that they require significant fine tuning and offer unsatisfactory performance. We thus propose an efficient and effective transfer-learning based approach for unsupervised anomaly detection. Our method employs a deep convolutional neural network trained on ImageNet as a feature extractor, and subsequently feeds the learned feature representations into an existing unsupervised anomaly detection method. We show that our approach significantly outperforms baselines on two natural image datasets and two retinal fundus image datasets, all with minimal fine-tuning. We further show the ability to leverage very small numbers of labelled anomalies to improve performance. Our work establishes a strong unsupervised baseline for image-based anomaly detection, alongside a flexible and scalable approach for screening applications.

In the literature, the task of dysarthric speech intelligibility assessment has been approached through development of different low-level feature representations, subspace modeling, phone confidence estimation or measurement of automatic speech recognition system accuracy. This paper proposes a novel approach where the intelligibility is estimated as the percentage of correct words uttered by a speaker with dysarthria by matching and verifying utterances of the speaker with dysarthria against control speakers’ utterances in phone posterior feature space and broad phonetic posterior feature space. Experimental validation of the proposed approach on the UA-Speech database, with posterior feature estimators trained on the data from auxiliary domain and language, obtained a best Pearson's correlation coefficient ( $r$ ) of 0.950 and Spearman's correlation coefficient ( $\rho$ ) of 0.957. Furthermore, replacing control speakers’ speech with speech synthesized by a neural text-to-speech system obtained a best $r$ of 0.931 and $\rho$ of 0.961.