Multi-scale Dictionary Research Articles

In-Vehicle Network Intrusion Detection System Using Convolutional Neural Network and Multi-Scale Histograms

Cybersecurity in modern vehicles has received increased attention from the research community in recent years. Intrusion Detection Systems (IDSs) are one of the techniques used to detect and mitigate cybersecurity risks. This paper proposes a novel implementation of an IDS for in-vehicle security networks based on the concept of multi-scale histograms, which capture the frequencies of message identifiers in CAN-bus in-vehicle networks. In comparison to existing approaches in the literature based on a single histogram, the proposed approach widens the informative context used by the IDS for traffic analysis by taking into consideration sequences of two and three CAN-bus messages to create multi-scale dictionaries. The histograms are created from windows of in-vehicle network traffic. A preliminary multi-scale histogram model is created using only legitimate traffic. Against this model, the IDS performs traffic analysis to create a feature space based on the correlation of the histograms. Then, the created feature space is given in input to a Convolutional Neural Network (CNN) for the identification of the windows of traffic where the attack is present. The proposed approach has been evaluated on two different public data sets achieving a very competitive performance in comparison to the literature.

Rain Removal of Single Image Based on Directional Gradient Priors

Images taken on rainy days often lose a significant amount of detailed information owing to the coverage of rain streaks, which interfere with the recognition and detection of the intelligent vision systems. It is, therefore, extremely important to recover clean rain-free images from the rain images. In this paper, we propose a rain removal method based on directional gradient priors, which aims to retain the structural information of the original rain image to the greatest extent possible while removing the rain streaks. First, to solve the problem of residual rain streaks, on the basis of the sparse convolutional coding model, two directional gradient regularization terms are proposed to constrain the direction information of the rain stripe. Then, for the rain layer coding in the directional gradient prior terms, a multi-scale dictionary is designed for convolutional sparse coding to detect rain stripes of different widths. Finally, to obtain a more accurate solution, the alternating direction method of multipliers (ADMM) is used to update the multi-scale dictionary and coding coefficients alternately to obtain a rainless image with rich details. Finally, experiments verify that the proposed algorithm achieves good results both subjectively and objectively.

A sparse representation strategy to eliminate pseudo-HFO events from intracranial EEG for seizure onset zone localization

Objective. High-frequency oscillations (HFOs) are considered a biomarker of the epileptogenic zone in intracranial EEG recordings. However, automated HFO detectors confound true oscillations with spurious events caused by the presence of artifacts. Approach. We hypothesized that, unlike pseudo-HFOs with sharp transients or arbitrary shapes, real HFOs have a signal characteristic that can be represented using a small number of oscillatory bases. Based on this hypothesis using a sparse representation framework, this study introduces a new classification approach to distinguish true HFOs from the pseudo-events that mislead seizure onset zone (SOZ) localization. Moreover, we further classified the HFOs into ripples and fast ripples by introducing an adaptive reconstruction scheme using sparse representation. By visualizing the raw waveforms and time-frequency representation of events recorded from 16 patients, three experts labeled 6400 candidate events that passed an initial amplitude-threshold-based HFO detector. We formed a redundant analytical multiscale dictionary built from smooth oscillatory Gabor atoms and represented each event with orthogonal matching pursuit by using a small number of dictionary elements. We used the approximation error and residual signal at each iteration to extract features that can distinguish the HFOs from any type of artifact regardless of their corresponding source. We validated our model on sixteen subjects with thirty minutes of continuous interictal intracranial EEG recording from each. Main results. We showed that the accuracy of SOZ detection after applying our method was significantly improved. In particular, we achieved a 96.65% classification accuracy in labeled events and a 17.57% improvement in SOZ detection on continuous data. Our sparse representation framework can also distinguish between ripples and fast ripples. Significance. We show that by using a sparse representation approach we can remove the pseudo-HFOs from the pool of events and improve the reliability of detected HFOs in large data sets and minimize manual artifact elimination.

Learning Dual Memory Dictionaries for Blind Face Restoration.

Blind face restoration is a challenging task due to the unknown, unsynthesizable and complex degradation, yet is valuable in many practical applications. To improve the performance of blind face restoration, recent works mainly treat the two aspects, i.e., generic and specific restoration, separately. In particular, generic restoration attempts to restore the results through general facial structure prior, while on the one hand, cannot generalize to real-world degraded observations due to the limited capability of direct CNNs' mappings in learning blind restoration, and on the other hand, fails to exploit the identity-specific details. On the contrary, specific restoration aims to incorporate the identity features from the reference of the same identity, in which the requirement of proper reference severely limits the application scenarios. Generally, it is a challenging and intractable task to improve the photo-realistic performance of blind restoration and adaptively handle the generic and specific restoration scenarios with a single unified model. Instead of implicitly learning the mapping from a low-quality image to its high-quality counterpart, this paper suggests a DMDNet by explicitly memorizing the generic and specific features through dual dictionaries. First, the generic dictionary learns the general facial priors from high-quality images of any identity, while the specific dictionary stores the identity-belonging features for each person individually. Second, to handle the degraded input with or without specific reference, dictionary transform module is suggested to read the relevant details from the dual dictionaries which are subsequently fused into the input features. Finally, multi-scale dictionaries are leveraged to benefit the coarse-to-fine restoration. The whole framework including the generic and specific dictionaries is optimized in an end-to-end manner and can be flexibly plugged into different application scenarios. Moreover, a new high-quality dataset, termed CelebRef-HQ, is constructed to promote the exploration of specific face restoration in the high-resolution space. Experimental results demonstrate that the proposed DMDNet performs favorably against the state of the arts in both quantitative and qualitative evaluation, and generates more photo-realistic results on the real-world low-quality images. The codes, models and the CelebRef-HQ dataset will be publicly available at https://github.com/csxmli2016/DMDNet.

Learning Multiscale Convolutional Dictionaries for Image Reconstruction

Convolutional neural networks (CNNs) have been tremendously successful in solving imaging inverse problems. To understand their success, an effective strategy is to construct simpler and mathematically more tractable convolutional sparse coding (CSC) models that share essential ingredients with CNNs. Existing CSC methods, however, underperform leading CNNs in challenging inverse problems. We hypothesize that the performance gap may be attributed in part to how they process images at different spatial scales: While many CNNs use multiscale feature representations, existing CSC models mostly rely on single-scale dictionaries. To close the performance gap, we thus propose a multiscale convolutional dictionary structure. The proposed dictionary structure is derived from the U-Net, arguably the most versatile and widely used CNN for image-to-image learning problems. We show that incorporating the proposed multiscale dictionary in an otherwise standard CSC framework yields performance competitive with state-of-the-art CNNs across a range of challenging inverse problems including CT and MRI reconstruction. Our work thus demonstrates the effectiveness and scalability of the multiscale CSC approach in solving challenging inverse problems.

Patchwise dictionary learning for video forest fire smoke detection in wavelet domain

Fire smoke detection in forest faces much more challenges than in local areas or indoors. Conventional fire smoke methods using static features extracted form single image directly cannot handle some disturbed factors exist in forest, such as shaking trees and fog/haze. On the other hand, it is difficult to capture the dynamic features of fire smoke while they are vital for fire smoke detection. This paper proposes an effective method of extracting features based on sequential images for smoke detection, quasi-dynamic features are extracted using on pixel-block arranged image sets. First, consecutive frames are rearranged into a new big image based on a pixel-block rule, so that the new image still keep the whole structure of original image on block level, and the change of different pixels in the same block, which comes from the same position of each frame, can reflect the dynamic features by special process. Second, to process the new arranged big image, driven by single image wavelet transform for denoting local signal characteristics, Dual Tree-Complex Wavelet Transform is utilized to decompose the new big image into eight images with the same size of the source image after several level-decomposition, including four directions (HH, HL, LH, LL) with two decomposed images in each direction, so that the changes from smoke in different orientation can be extracted and depicted in these subbands. Third, a multi-scale dictionary learning method is proposed to learn dictionaries of each group images. Furthermore, elastic net is introduced for sparse coding and feature coefficients extraction. All operations above contribute to a significant improvement for smoke detection in this paper. Extensive experiments are performed to validate the effectiveness of the proposed approach.

Variational Multiscale Nonparametric Regression: Algorithms and Implementation

Many modern statistically efficient methods come with tremendous computational challenges, often leading to large-scale optimisation problems. In this work, we examine such computational issues for recently developed estimation methods in nonparametric regression with a specific view on image denoising. We consider in particular certain variational multiscale estimators which are statistically optimal in minimax sense, yet computationally intensive. Such an estimator is computed as the minimiser of a smoothness functional (e.g., TV norm) over the class of all estimators such that none of its coefficients with respect to a given multiscale dictionary is statistically significant. The so obtained multiscale Nemirowski-Dantzig estimator (MIND) can incorporate any convex smoothness functional and combine it with a proper dictionary including wavelets, curvelets and shearlets. The computation of MIND in general requires to solve a high-dimensional constrained convex optimisation problem with a specific structure of the constraints induced by the statistical multiscale testing criterion. To solve this explicitly, we discuss three different algorithmic approaches: the Chambolle-Pock, ADMM and semismooth Newton algorithms. Algorithmic details and an explicit implementation is presented and the solutions are then compared numerically in a simulation study and on various test images. We thereby recommend the Chambolle-Pock algorithm in most cases for its fast convergence. We stress that our analysis can also be transferred to signal recovery and other denoising problems to recover more general objects whenever it is possible to borrow statistical strength from data patches of similar object structure.

Faster Multiscale Dictionary Learning Method With Adaptive Parameter Estimation for Fault Diagnosis of Traction Motor Bearings

Traction motor bearings are crucial components to guarantee stable operation. Thus, it is significant to monitor the bearing condition. Dictionary learning is a powerful method to extract the characteristic condition. Compared with ordinary dictionary learning, the multiscale dictionary learning method applied to transform coefficients performs well in extracting the impact signals and takes less learning time. However, it spends more time tuning parameters to make the algorithm more efficient, especially in practical industrial applications. Hence, a faster adaptive parameter multiscale dictionary learning (Faster AP-MSDL) method is proposed in this article which adaptively chooses the scales of learning and estimates the parameters of sparse coding in dictionary learning simultaneously, and it possesses two core traits including less learning time and adaptive parameter estimation, which make this method more suitable to practical industrial application. Finally, a simulation experiment and two practical industrial experiments including a bearing run-to-failure test and a fault experiment of traction motor rolling bearing are conducted using our proposed method to verify adaptive ability and superiority. The superiority of the proposed method is also verified by comparing with other state-of-the-art methods.

Multi-Scale Structured Dictionary Learning for 3-D Point Cloud Attribute Compression

Existing 3-D point cloud attribute compression methods are rigid for complex signals defined in spatially irregular domain, as they solely depend on geometric information or adopt pre-defined analytic bases for transform. In this article, we propose a multi-scale structured dictionary learning algorithm for 3-D point cloud attribute compression. The proposed multi-scale dictionary leverages hierarchical sparsity within signals by adapting atoms in an arborescent structure. It supports hierarchical sparse transform for encoding voxelized point cloud attributes with progressive refinement of high-frequency details from coarser to finer scales. To represent spatial blocks with varying amounts of occupied voxels, a non-uniform binary weight matrix is incorporated in dictionary learning to characterize the dimensional irregularity of signals. To guarantee optimal approximation with improved efficiency, alternating optimizations including Alternating Direction Method of Multipliers and Gauss-Seidel iterations are adopted for the weighted mixed $\ell _{2}/\ell _{1}$ -minimization. Furthermore, we develop a lossy attribute compression framework by integrating sparse transform based prediction with dedicated quantization and entropy coding routines. Experimental results demonstrate that the proposed framework outperforms the state-of-the-art transform-based coding methods and is competitive with the most recent MPEG PCC test models in point cloud attribute compression.

Fingerprint enhancement using multi‐scale classification dictionaries with reduced dimensionality

In order to improve the quality of fingerprint with a large noise, this study proposes a fingerprint enhancement method by using a sparse representation of learned multi-scale classification dictionaries with reduced dimensionality. The multi-scale dictionary is used to balance the contradiction between the accuracy and the anti-noise ability, which is an ideal solution to reconcile the demands of enhancement quality and computational performance. The principal component analysis is applied in the authors’ technique for dimension reduction of multi-scale classification dictionaries. Under the quality grading scheme and multi-scale composite windows, the fingerprint patches are enhanced by using a sparse representation of learned multi-scale classification dictionaries with reduced dimensionality according to their priorities. In addition, the multi-scale composite windows help the more high-quality spectra diffuse into the low-quality fingerprint patches and this can greatly improve the spectra quality of them. Experimental results and comparisons on FVC 2000 and FVC 2004 databases are reported. And it shows that the proposed method yields better results in terms of the robustness of fingerprint enhancement as compared with the latest techniques. Moreover, the results show that the proposed algorithm can obtain better identification performance.

Salient feature multimodal image fusion with a joint sparse model and multiscale dictionary learning

A multimodal image fusion method based on the joint sparse model (JSM), multiscale dictionary learning, and a structural similarity index (SSIM) is presented. As an effective signal representation technique, JSM is derived from distributed compressed sensing and has been successfully employed in many image-processing applications such as image classification and fusion. The highly redundant single dictionary always has difficulty satisfying the correlations between images in traditional JSM-based image fusion. Therefore, the proposed fusion model learns a more compact multiscale dictionary to effectively combine the multiscale analysis used in nonsubsampled contourlet transformation with the single-scale joint sparse representation used in image domains to solve the issues of single-scale sparse fusion and to improve fusion quality. The experimental results demonstrate that the proposed fusion method obtains the state-of-the-art performances in terms of both subjective visual quality and objective metrics, especially when fusing multimodal images.

Axis-guided patch based accurate segmentation for pathological vessels using adaptive weight sparse representation

Background and objectiveOver decades of research and development in vessel segmentation, accurate and reliable methods targeting at pathological vessels are few. Addressing this challenge, we try to delineate the vessel boundary accurately with the presence of pathologies. MethodsA novel segmentation framework is presented in this work, a vessel axis tracking algorithm + a patch-based sparse representation. The patch-based algorithm is navigated by the vessel axis tracking algorithm. Within the training process, multi-scale training samples have been used, which has the potential to be both physiologically accurate and computationally effective. The redundant information embedded in the multi-scale training samples is employed to delineate the pathological vessel accurately. To further reduce the computational burden caused by the patch-based sparse representation, a multi-scale dictionary has been generated, and adaptive weights have been assigned to the scale-specific sub-dictionary atoms, for computational efficiency. ResultsOur method is evaluated by comparing with two state-of-the-art methods, on synthetic complex-structured datasets and real clinical datasets. The performance of our method is promising over the evaluation, since the overlap ratios of our method are high over all the datasets, around 91%, much better than two state-of-the-art methods.

Tool Condition Monitoring With Multiscale Discriminant Sparse Decomposition

Sparse decomposition has been successfully applied in manufacturing process condition monitoring. In the sparse decomposition-based approaches, the representation dictionary plays an important role. However, most current dictionary learning based methods do not consider the atoms corresponding to the specific states in the dictionary, and the coding coefficients lack discrimination between different states. This study develops an approach capable of multiscale dictionary learning for tool condition monitoring (TCM) and improving the discrimination power by maximizing the between-class and within-class coding coefficients. In addition, through the use of the discriminant distance without signal reconstruction, it enhances the state discrimination and increases the speed of online monitoring. The experimental results have validated the effectiveness of the approach with applications to micromilling TCM.

Multi-scale dictionary learning for ocular artifact reduction from single-channel electroencephalograms

This paper addresses two issues toward practical use of wearable electroencephalogram (EEG) measurement devices. Ocular (eye movement and blink) artifacts often contaminate EEGs and deteriorate the performance of EEG-based brain–computer interfaces (BCIs). Although wearable consumer-grade EEG devices with single electrode allow users to operate BCIs conveniently in daily lives, it remains a challenging issue to attenuate ocular artifacts from single-channel measurements without spatial information. Existing ocular artifact reduction methods are, however, not simple enough for single-channel EEG data in the sense that they require an additional reference channel and/or pre-processing for artifact segment detection. Another issue is how to assess the performance of artifact reduction; the existing studies have used their own datasets that are not accessible from other researchers. Then, this paper makes two major contributions. (1) This paper proposes a novel ocular artifact reduction method, multi-scale dictionary learning (MSDL), which operates under single-channel measurements and without artifact segment detection. (2) We also develop a semi-simulation setting for quantitative evaluation with a publicly available EEG dataset. In particular, we employed BCI Competition IV Dataset 2a, on which the proposed method was compared with state-of-art methods. The proposed technique showed the best performance for recovering artifact-reduced waveforms from single-channel data compared to the other artifact reduction methods. The Matlab scripts for semi-simulation data generation and single-channel artifact reduction are available on GitHub.

Fabric Defect Detection and Classifier via Multi-Scale Dictionary Learning and an Adaptive Differential Evolution Optimized Regularization Extreme Learning Machine

To develop an automatic detection and classifier model for fabric defects, a novel detection and classifier technique based on multi-scale dictionary learning and the adaptive differential evolution algorithm optimised regularisation extreme learning machine (ADE-RELM) is proposed. Firstly in order to speed up dictionary updating under the condition of guaranteeing dictionary sparseness, k-means singular value decomposition (KSVD) dictionary learning is used. Then multi-scale KSVD dictionary learning is presented to extract texture features of textile images more accurately. Finally a unique ADE-RELM is designed to build a defect classifier model. In the training ADE-RELM classifier stage, a self-adaptive mutation operator is used to solve the parameter setting problem of the original differential evolution algorithm, then the adaptive differential evolution algorithm is utilised to calculate the optimal input weights and hidden bias of RELM. The method proposed is committed to detecting common defects like broken warp, broken weft, oil, and the declining warp of grey-level and pure colour fabrics. Experimental results show that compared with the traditional Gabor filter method, morphological operation and local binary pattern, the method proposed in this paper can locate defects precisely and achieve high detection efficiency.

A weighted multi-scale dictionary learning model and its applications on bearing fault diagnosis

Extracting impulsive information under strong background noise and harmonic interference is a challenging problem for bearing fault diagnosis. Multi-scale transforms have achieved great success in extracting impulsive feature information, however, how to choose a suitable transform is a difficult problem, especially in the case of strong noise interference. Therefore, dictionary learning methods have attracted more and more attention in recent years. A weighted multi-scale dictionary learning model (WMSDL) is proposed in this paper which integrates the multi-scale transform and fault information into a unified dictionary learning model and it successfully overcomes four disadvantages of traditional dictionary learning algorithms including lacking the multi-scale property; restricting training samples to local patches; being sensitive to strong harmonic interference; suffering from high computational complexity. Moreover, algorithmic derivation, computational complexity and parameter selection are discussed. Finally, The effectiveness of the proposed method is verified by both the numerical simulations and experiments. Comparisons with other state-of-the-art methods further demonstrate the superiority of the proposed method.

Data‐driven multiscale sparse representation for bearing fault diagnosis in wind turbine

AbstractWith the increase of the wind turbine capacity, failures occur on the drivetrain of wind turbines frequently. Since faults of bearings in the wind turbine can lead to long downtime and even casualties, fault diagnosis of the drivetrain is very important to reduce the maintenance cost of the wind turbine and improve economic efficiency. However, the traditional diagnosis methods have difficulty in extracting the impulsive components from the vibration signal of the wind turbine because of heavy background noise and harmonic interference. In this paper, we propose a novel method based on data‐driven multiscale dictionary construction. Firstly, we achieve the useful atom through training the K‐means singular value decomposition (K‐SVD) model with a standard signal. Secondly, we deform the chosen atom into different shapes and construct the final dictionary. Thirdly, the constructed dictionary is used to sparsely represent the vibration signal, and orthogonal matching pursuit (OMP) is performed to extract the impulsive component. The proposed method is robust to harmonic interference and heavy background noise. Moreover, the effectiveness of the proposed method is validated by numerical simulation and two experimental cases including the bearing fault of the wind turbine generator in the field test. The overall results indicate that compared with traditional methods, the proposed method is able to extract the fault characteristics from the measured signals more efficiently.

Super-resolution reconstruction algorithm based on multi-scale non-local dictionary training for interested area of UAV

ABSTRACTThe multi-frame images taken by the unmanned aerial vehicle (UAV) in the complex flight environment will be easily affected by the jitter, noise and exposure, which result in the decline of detection rate of the interested regional target. Considering this problem, a super-resolution reconstruction algorithm for aerial multi-frame images of UAV is proposed by using multi-scale non-local dictionary training method based on compressive sensing. Experimental results prove the validity and accuracy of the proposed algorithm.

An Efficient Method to Learn Overcomplete Multi-Scale Dictionaries of ECG Signals

The electrocardiogram (ECG) was the first biomedical signal for which digital signal processing techniques were extensively applied. By its own nature, the ECG is typically a sparse signal, composed of regular activations (QRS complexes and other waveforms, such as the P and T waves) and periods of inactivity (corresponding to isoelectric intervals, such as the PQ or ST segments), plus noise and interferences. In this work, we describe an efficient method to construct an overcomplete and multi-scale dictionary for sparse ECG representation using waveforms recorded from real-world patients. Unlike most existing methods (which require multiple alternative iterations of the dictionary learning and sparse representation stages), the proposed approach learns the dictionary first, and then applies a fast sparse inference algorithm to model the signal using the constructed dictionary. As a result, our method is much more efficient from a computational point of view than other existing algorithms, thus becoming amenable to dealing with long recordings from multiple patients. Regarding the dictionary construction, we located first all the QRS complexes in the training database, then we computed a single average waveform per patient, and finally we selected the most representative waveforms (using a correlation-based approach) as the basic atoms that were resampled to construct the multi-scale dictionary. Simulations on real-world records from Physionet’s PTB database show the good performance of the proposed approach.

Finding GEMS: Multi-Scale Dictionaries For High-Dimensional Graph Signals

Modern data introduces new challenges to classic signal processing approaches, leading to a growing interest in the field of graph signal processing. A powerful and well established model for real world signals in various domains is sparse representation over a dictionary, combined with the ability to train the dictionary from signal examples. This model has been successfully applied to graph signals as well by integrating the underlying graph topology into the learned dictionary. Nonetheless, dictionary learning methods for graph signals are typically restricted to small dimensions due to the computational constraints that the dictionary learning problem entails, and due to the direct use of the graph Laplacian matrix. In this paper, we propose a dictionary learning algorithm that applies to a broader class of graph signals, and is capable of handling much higher dimensional data. We incorporate the underlying graph topology both implicitly, by forcing the learned dictionary atoms to be sparse combinations of graph-wavelet functions, and explicitly, by adding direct graph constraints to promote smoothness in both the feature and manifold domains. The resulting atoms are thus adapted to the data of interest while adhering to the underlying graph structure and possessing a desired multi-scale property. Experimental results on several datasets, representing both synthetic and real network data of different nature, demonstrate the effectiveness of the proposed algorithm for graph signal processing even in high dimensions.