Search

Adversarial training has been considered to be one of the most effective strategies to defend against adversarial attacks. Most existing adversarial training methods have shown a trade-off between training cost and robustness. This paper explores a new optimization direction from training data to reduce the computational cost of adversarial training without scarifying robustness. First, we show that some adversarial examples are less important, meaning that removing them does not hurt the robustness. Second, we propose a method to identify insignificant adversarial examples at a minimal cost. Third, we demonstrate that our approach can be integrated with other adversarial training frameworks with few modifications. The experimental results show that combined with previous works, our approach not only reduces about 20% of computational cost on the CIFAR10 and CIFAR100 datasets but also improves about 1.5% natural accuracy. With less computational cost, it achieves 58.22%, 30.68%, and 41.92% robust accuracy on CIFAR10, CIFAR100, and ImageNet datasets respectively, which are higher than those of the original methods. • Analyze the effectiveness of adversarial examples for adversarial training. • Design a method to remove inefficient adversarial examples but maintain the accuracy. • Design an algorithm to detect inefficient adversarial examples without extra cost. • Conduct experiments to demonstrate the compatibility with other frameworks. [ABSTRACT FROM AUTHOR]

Cluster analysis is one of the most important research issues in data mining and machine learning. To date, numerous clustering algorithms have been proposed to tackle the fixed-length vector data. In many real applications, we need to detect clusters from a set of discrete sequences in which each sequence is an ordered list of items. Due to the sequential and discrete nature, the discrete sequence clustering problem is more challenging and most of existing vector data clustering algorithms cannot be directly employed. In this paper, we present a stochastic algorithm for clustering discrete sequences. Our method first quickly generates a set of random partitions over the sequential data set and then merges these random clustering results via weighted graph construction and partition. We perform extensive empirical comparisons on real data sets to show that our method is comparable to those state-of-the-art clustering algorithms with respect to both accuracy and efficiency. • A randomized algorithm is proposed to solve the discrete sequence clustering issue. • Our algorithm provides high success rate for sequence linkage in clusters. • Experimental results demonstrate the feasibility and effectiveness of our method. [ABSTRACT FROM AUTHOR]

This paper investigates algorithms for matrix factorization when some or many components are missing, a problem that arises frequently in computer vision and pattern recognition. We demonstrate that the Jacobian used in the damped Wiberg (DW) method is exactly the same as that of Kaufman's simplified variable projection (VP) algorithm. Our analysis provides a novel perspective on the efficiency of VP algorithms by improving the strong convexity of the approximate function. To enhance numerical stability, we set a lower bound on the damping parameter instead of adding a null space like the DW algorithm. Another challenge of low-rank matrix decomposition with missing data is the existence of many sharp local minima, which are often distributed in narrow valleys of the landscape of objection functions. Falling into such minima tends to result in poor reconstruction results. To address this issue, we design a non-monotonic VP algorithm, which can facilitate the algorithm to escape from sharp minima and converge to flatter minima. Numerical experiments confirm the effectiveness and efficiency of the proposed nonmonotone VP algorithm. • We provide a new perspective to examine the superiority of the VP algorithm from a convexity point of view. • We propose a nonmonotone VP algorithm, which can facilitate the algorithm to escape from sharp minima. • We conduct extensive experiments to validate the effectiveness of the proposed nonmonotone algorithm. [ABSTRACT FROM AUTHOR]

Wide-angle video frames typically contain more information, but they also exhibit distortion that degrades the visual quality, especially at the edges. To eliminate this distortion from videos, we propose a self-supervised iterative optimization method in this paper. Specifically, we construct a motion parameter estimation model utilizing two consecutive distorted frames, where motion parameters comprise affine transform and distortion parameters. We apply the Gauss–Newton algorithm to minimize the sum-of-squares error between frames and update parameters. Treating inter-frame motion as undistort-affine-distort transformations, frame alignment is achieved by continuously adjusting transform parameters. Ultimately, frames are corrected using the converged parameters. We generated a synthetic dataset with various distortion parameters for evaluation. Experiments demonstrate superior performance versus state-of-the-art methods on synthetic and real wide-angle videos. Our algorithm also achieves higher parameter estimation accuracy without sacrificing efficiency. • proposes a algorithm for video distortion correction based on iterative optimization. • the distort transform and undistort transform are derived separately and the range of parameters and the experimental error are given. • Based on this, we transform the distortion correction into a parameter estimation problem in image alignment. • optimize the parameter estimation problem using Gauss-Newton algorithm. • Experimental results show that the proposed algorithms outperform some state-of-art algorithms in accuracy, without reducing robustness and efficiency. [ABSTRACT FROM AUTHOR]

In this paper, we investigate L 0 -regularized logistic regression models, and design two fast and efficient algorithms for high-dimensional correlated data and massive data, respectively. Our first algorithm, the Variable Sorted Active Set (VSAS) algorithm, is based on the local quadratic approximation of the KKT conditions for L 0 -penalized maximum log-likelihood function in high-dimensional correlated data. We establish an L ∞ error upper bound for the estimator obtained by the VSAS algorithm and prove its optimal convergence rate. Moreover, when the target signal exceeds the detectable level, the estimator obtained by the VSAS algorithm can achieve the oracle estimator with high probability. Our second algorithm, Communication Effective Variable Sorted Active Set (CEVSAS), aims to solve high-dimensional and large-sample L 0 -regularized logistic regression models by reduce computational and communication costs, while maintaining estimation efficiency. Finally, simulations and real data demonstrate the effectiveness of our proposed VSAS and CEVSAS algorithms. • We develop a fast VSAS algorithm for L 0 regularized logistic regression involving high-dimensional correlated data. • An L ∞ upper bound is established on the estimation error of the solution sequence generated by VSAS algorithm. • The CEVSAS altorithm is proposed to reduce computational costs and ensure higher estimation efficiency. • We propose two adaptive versions of the VSAS and CEVSAS algorithms for selecting the optimal tuning parameters. [ABSTRACT FROM AUTHOR]

• We propose an improved differential evolution (DE) algorithm by composite binary crossover and Pareto refining operations to the DE. • DE solves the quantifying fraudulent transactions problem by minimizing the selected feature cost while maximizing the quantification ability. • We show that the proposed DE outperforms four other variants of DE and five state-of-the-art EAs on various credit datasets. • The results are also statically validated through the Friedman test and Wilcoxon post-hoc analysis. [Display omitted] Identification of fraudulent credit card transactions is a complex problem mainly due to the following factors: 1) The relative behavior of customers and fraudsters may alter over time. 2) The ratio of legitimate to fraudulent transactions is highly imbalanced, and 3) Investigators examine a small segment of transactions in a reasonable time frame. Researchers have proposed various algorithms to identify potential fraud in a new incoming transaction. However, these approaches require significant human investigator effort and are sometimes misleading. To address this issue, this paper proposes an improved multiobjective differential evolution (DE) algorithm to estimate the distribution of fraudulent transactions in a set of new incoming transactions, referred to as quantifying fraudulent transactions. Our paper has three major novelties. First, we present the problem formulation of cost-based feature selection with maximum quantification ability. Second, we improve the DE by applying effective trial vector generation algorithms to the random control parameter settings to exploit the advantage of individual DE variants. Third, we develop the maximum-relevancy-minimum-redundancy-based Pareto refining operator to enhance the self-learning ability of individuals in Pareto solutions. We compare our approach against four other modifications of DE and five state-of-the-art evolutionary algorithms on real-time credit datasets in streaming and non-streaming frameworks using hyper-volume, two-set coverage, and spread performance metrics. [ABSTRACT FROM AUTHOR]

Highlights • This paper proposes a novel connected components labeling (CCL) approach, a gamma-signal-regulated approach, for higher efficiency. • A new and more efficient block-based two-scan CCL algorithm, Eight-Connected Gamma-Signal-regulated (ECGS) algorithm, is proposed. • Experiment results have demonstrated that ECGS can outperform current state-of-the-art CCL algorithms for a number of digital images. Abstract This paper proposes a novel connected components labeling (CCL) approach that introduces a gamma signal to record certain mask pixels' values to eliminate duplicated pixel checking and regulate the labeling process for higher efficiency. A new block-based two-scan CCL algorithm, Eight-Connected Gamma-Signal-regulated (ECGS) algorithm, is designed and developed by applying this approach to evaluate a block of 2 × 2 pixels (with just 6 mask pixels) in each iteration such that the total number of operations is considerably reduced and the labeling efficiency is significantly improved. The experiments conducted on a public benchmark, YACCLAB (Yet Another Connected Components Labeling Benchmark), have demonstrated that the proposed ECGS algorithm can outperform current state-of-the-art CCL algorithms for a number of digital images. [ABSTRACT FROM AUTHOR]

Highlights • We find an observation important to the LSN subspace segmentation. • We provide the theoretical support of the mentioned observation. • Our paper is the first to adopt infinite norm in subspace segmentation. • The computational complexity of our ADM algorithm is lower than O (n 3). • Our method achieve the state-of-the-art results in extensive experiments. Abstract Spectral-clustering based methods have recently attracted considerable attention in the field of subspace segmentation. The approximately block-diagonal graphs achieved by this kind of methods usually contain some noise, i.e., nonzero elements in the off-diagonal region, due to outlier contamination or complex intrinsic structure of the dataset. In the experiment of most previous work, the number of the subspaces is often no more than 10. In this situation, this kind of noise almost has no influence on the segmentation results. However, the segmentation performance could be negatively affected by the noise when the number of subspaces is large, which is quite common in the real-world applications. In this paper, we address the problem of LSN subspace segmentation, i.e., large subspace number subspace segmentation. We first show that the approximately block-diagonal graph with the smaller difference in its diagonal blocks will be more robust to the off-diagonal noise mentioned above. Then, by using the infinity norm to control the bound of the difference in the diagonal blocks, we propose infinity norm minimization for LSN subspace segmentation. Experimental results demonstrate the effectiveness of our method. [ABSTRACT FROM AUTHOR]

Highlights • We propose a linear classifier design method in the weight space based on the observation that a training sample represents a plane in the weight space. • We proposed search algorithms to find the optimal subspace, though the optimal solution may not be guaranteed since the full search cannot be used. • Experimental results show that the proposed classifier performed better than or equivalent to existing linear classifiers (linear SVM, LDA). Abstract In this paper, we propose a linear classifier design method in the weight space. A linear classifier is completely determined by a weight vector. To design a linear classifier is equivalent to finding a weight vector. When there are a number of training samples, each training sample represents a plane in the weight space. On one side of the plane, the training sample is correctly classified while it is incorrectly classified on the other side. Thus, finding the optimal linear classifier can be formulated as finding a subspace that provides the best classification accuracy. In this paper, we have developed an algorithm to find such optimal subspaces in the weight space. Although we could not find the optimal linear classifier due to prohibitive computational complexity, the proposed design method produced noticeable improvements in some cases. Experimental results show that the proposed linear classifier performed better than or equivalent to existing linear classifiers. [ABSTRACT FROM AUTHOR]

Highlights • First system for Simultaneous Localization and Mapping in large environments from Squared Planar Markers. • Print markers with a regular printer and place them in environment. Use a camera an our system finds the camera position while estimating the pose of the markers at the same time. • Real time operation on CPUs (150 Hz) with a single execution thread. Abstract SLAM is generally addressed using natural landmarks such as keypoints or texture, but it poses some limitations, such as the need for enough textured environments and high computational demands. In some cases, it is preferable sacrificing the flexibility of such methods for an increase in speed and robustness by using artificial landmarks. The recent work [1] proposes an off-line method to obtain a map of squared planar markers in large indoor environments. By freely distributing a set of markers printed on a piece of paper, the method estimates the marker poses from a set of images, given that at least two markers are visible in each image. Afterwards, camera localization can be done, in the correct scale. However, an off-line process has several limitations. First, errors can not be detected until the whole process is finished, e.g., an insufficient number of markers in the scene or markers not properly spotted in the capture stage. Second, the method is not incremental, so, in case of requiring the expansion of the map, it is necessary to repeat the whole process from start. Finally, the method can not be employed in real-time systems with limited computational resources such as mobile robots or UAVs. To solve these limitations, this work proposes a real-time solution to the problems of simultaneously localizing the camera and building a map of planar markers. This paper contributes with a number of solutions to the problems arising when solving SLAM from squared planar markers, coining the term SPM-SLAM. The experiments carried out show that our method can be more robust, precise and fast, than visual SLAM methods based on keypoints or texture. [ABSTRACT FROM AUTHOR]

This paper proposes a new method of personal identification that introduces the analysis of lip prints. In spite of its important role in forensic and biometric applications, the results of previous investigations into lip prints are scanty. This is mainly due to the difficulties that accompany any analysis of the lips: lips are very flexible and pliable, and successive lip print impressions – even those obtained from the same person – may significantly differ from one other. Our article’s principal contribution is a proposal for a new personal identification methodology and application that uses, for the first time, a strategy for biometric and forensic analysis of lip print structure. As a result of our analysis we propose a lip print pattern for each individual. This pattern contains only such furrows that occur on the greatest number of lip prints obtained from the same person, where these furrows’ locations and inclinations remain similar across the lip prints obtained. It should be noted that in our approach, instead of lip photos we employ lip prints, such as can be obtained at a crime scene. It is worth noticing also that we propose a new method of personal identification where, instead of popular machine learning methods,the furrow-analysis of lip prints is introduced. Thus, no classifier learning phase is required. According to the authors’ convictions, based on reports in the literature, the proposed approach describes for the first time a strategy as to how lip print structures could be analyzed in biometric applications. The main novelty of this paper is its use of multiple lip print furrows from the same person to determine the most common lip furrow distribution in four different directions. On the basis of the pattern recognition of the lip furrows, the identity of the person will be determined. The proposed method’s effectiveness is 92.73%, determined using a database containing 350 lip prints. [ABSTRACT FROM AUTHOR]

This paper explores the problem of circle fitting for incomplete (partial arc) laser scanning point cloud data in the presence of outliers. In mobile laser scanning, data are commonly incomplete because of the orientation of the scanning unit to the surveying objects and the limited street-based positions. Also, multiple structures in the built environment often produce clustered outliers. To address these problems, this paper combines robust Principal Component Analysis (PCA) and robust regression with an efficient algebraic circle fitting method to develop two algorithms for circle fitting. Experimental efforts show that the proposed algorithms are statistically robust and can tolerate a high-percentage (exceeding 44%) of clustered outliers with insignificant error levels, while still achieving better shape recognition compared to existing competitive methods. For example, for a simulation of 1000 quarter circle datasets including 20% clustered outliers, RANSAC estimated the circle radius with a Mean Squared Error (MSE) of 172.10, whereas the proposed algorithms fit circles with an MSE of less than 0.42. The algorithms have potential in many areas including building information modeling, particle tracking, product quality control, arboreal assessment, and road asset monitoring. [ABSTRACT FROM AUTHOR]

• The first attempt to design a GEP-based codematrix generation algorithm. The individual structure of GEP allows the codematrix to be optimized in a much faster and more effective way. • The feature selection scheme is combined with the design of codematrix, to accomplish the feature selection and codematrix optimization simultaneously. • A greedy strategy-based adaptive optimization method to further improve the individual's quality to deal with hard classes by appending proper columns to the codematrix. The multi-class classification task is one of the most common tasks in machine learning. As a typical solution based on a partitioning strategy, Error-Correcting Output Codes (ECOC) can transform a multi-class classification problem into multiple binary classification problems. The key of ECOC is to construct an effective codematrix to represent a set of class decomposition schemes, which transforms a multiclass problem into a group of binary class problems. Consequently, the design of a fast and effective ECOC codematrix generation method is of great research significance and value for solving multi-class classification problems. In ECOC algorithms, the design of codematrix is treated as a combination problem between different code columns, in which the evolutionary algorithm shows a great advantage. Based on this consideration, the Gene Expression Programming (GEP) is applied to search for the codematrix with high performance because its expressive tree structure makes it well represent codematrcies for subsequent optimization operations. This paper proposes an adaptive ECOC algorithm based on Gene Expression Programming (GEP) and similarity measurement matrix, named GEP-ECOC. In our GEP, each individual represents a set of columns to form a random ECOC codematrix, which is optimized in the evolutionary process. Meanwhile, the crossover and mutation operations are modified to include a legality checking process to ensure that the generated codematrix satisfies the ECOC constraints. The GEP-based ECOC codematrix generation algorithm can quickly produce a codematrix with better performance, which ensures the efficiency of the algorithm to a certain extent. In addition, an adaptive algorithm based on a similarity measurement matrix is proposed to add new columns to the current codematrix, aiming to better handle hard classes. Our algorithm is compared with other algorithms on various data sets, and the experimental results confirm that our GEP-ECOC can balance the efficiency and performance of the algorithm and achieve higher performance. [ABSTRACT FROM AUTHOR]

Class imbalance problem commonly exists in multi-label classification (MLC) tasks. It has non-negligible impacts on the classifier performance and has drawn extensive attention in recent years. Borderline oversampling has been widely used in single-label learning as a competitive technique in dealing with class imbalance. Nevertheless, the borderline samples in multi-label data sets (MLDs) have not been studied. Hence, this paper deeply discussed the borderline samples in MLDs and found they have different neighboring relationships with class borders, which makes their roles different in the classifier training. For that, they are divided into two types named the self-borderline samples and the cross-borderline samples. Further, a novel MLDs resampling approach called Multi-Label Borderline Oversampling Technique (MLBOTE) is proposed for multi-label imbalanced learning. MLBOTE identifies three types of seed samples, including interior, self-borderline, and cross-borderline samples, and different oversampling mechanisms are designed for them, respectively. Meanwhile, it regards not only the minority classes but also the classes suffering from one-vs-rest imbalance as those in need of oversampling. Experiments on eight data sets with nine MLC algorithms and three base classifiers are carried out to compare MLBOTE with some state-of-art MLDs resampling techniques. The results show MLBOTE outperforms other methods in various scenarios. • A new borderline oversampling technique for multi-label imbalanced learning. • Defining self-borderline and cross-borderline samples in multi-label data sets. • Handling one-vs-rest imbalance in multi-label imbalanced learning. • Performing competitively in the experiments with C4.5, RBF kernel SVM and linear SVM. [ABSTRACT FROM AUTHOR]

Channel pruning is broadly recognized as an effective approach to obtain a small compact model through eliminating unimportant channels from a large cumbersome network. Contemporary methods typically perform iterative pruning procedure from the original over-parameterized model, which is both tedious and expensive especially when the pruning is aggressive. In this paper, we propose a simple yet effective channel pruning technique, termed network Pruning via rEsource rEalLocation (PEEL), to quickly produce a desired slim model with negligible cost. Specifically, PEEL first constructs a predefined backbone and then conducts resource reallocation on it to shift parameters from less informative layers to more important layers in one round, thus amplifying the positive effect of these informative layers. To demonstrate the effectiveness of PEEL , we perform extensive experiments on ImageNet with ResNet-18, ResNet-50, MobileNetV2, MobileNetV3-small and EfficientNet-B0. Experimental results show that structures uncovered by PEEL exhibit competitive performance with state-of-the-art pruning algorithms under various pruning settings. Encouraging results are also observed when applying PEEL to compress the semantic segmentation model. Our code is available at https://github.com/cardwing/Codes-for-PEEL. • Previous methods typically perform iterative pruning and are tedious and expensive. • A simple yet effective method is proposed to quickly produce a slim model with negligible cost. • Our method constructs a predefined backbone and conducts resource reallocation on it. • Resource reallocation shifts parameters from less informative layers to more important ones in one round. • Extensive experiments on ImageNet with various models validate the effectiveness. [ABSTRACT FROM AUTHOR]

This guest editorial introduces the twenty two papers accepted for this Special Issue on Articulated Motion and Deformable Objects (AMDO). They are grouped into four main categories within the field of AMDO: human motion analysis (action/gesture), human pose estimation, deformable shape segmentation, and face analysis. For each of the four topics, a survey of the recent developments in the field is presented. The accepted papers are briefly introduced in the context of this survey. They contribute novel methods, algorithms with improved performance as measured on benchmarking datasets, as well as two new datasets for hand action detection and human posture analysis. The special issue should be of high relevance to the reader interested in AMDO recognition and promote future research directions in the field. [ABSTRACT FROM AUTHOR]

Multiple instance learning (MIL) is a form of weakly supervised learning where training instances are arranged in sets, called bags, and a label is provided for the entire bag. This formulation is gaining interest because it naturally fits various problems and allows to leverage weakly labeled data. Consequently, it has been used in diverse application fields such as computer vision and document classification. However, learning from bags raises important challenges that are unique to MIL. This paper provides a comprehensive survey of the characteristics which define and differentiate the types of MIL problems. Until now, these problem characteristics have not been formally identified and described. As a result, the variations in performance of MIL algorithms from one data set to another are difficult to explain. In this paper, MIL problem characteristics are grouped into four broad categories: the composition of the bags, the types of data distribution, the ambiguity of instance labels, and the task to be performed. Methods specialized to address each category are reviewed. Then, the extent to which these characteristics manifest themselves in key MIL application areas are described. Finally, experiments are conducted to compare the performance of 16 state-of-the-art MIL methods on selected problem characteristics. This paper provides insight on how the problem characteristics affect MIL algorithms, recommendations for future benchmarking and promising avenues for research. Code is available on-line at https://github.com/macarbonneau/MILSurvey . [ABSTRACT FROM AUTHOR]

Ensemble methods have shown to be more effective than monolithic classifiers, in particular when diversity holds among their components. How to enforce diversity in classifier ensembles has received much attention from machine learning researchers, yielding a variety of different techniques and algorithms. In this paper, a novel algorithm for ensemble classifiers is proposed, in which ensemble components are trained with focus on different regions of the sample space. In so doing, diversity is mainly a consequence of the intention to limit the scope of base classifiers. The algorithm proposed in this paper shares roots with several ensemble paradigms, in particular with random forests, as it generates forests of decision trees as well. As decision trees are trained with focus on specific subsets of the sample space, the resulting ensemble is in fact a forest of “local” trees. Comparative experimental results highlight that, on average, these ensembles perform better than other relevant kinds of ensemble classifiers, including random forests. [ABSTRACT FROM AUTHOR]

The term “iris” refers to the highly textured annular portion of the human eye that is externally visible. An iris recognition system exploits the richness of these textural patterns to distinguish individuals. Iris recognition systems are being used in a number of human recognition applications such as access control, national ID schemes, border control, etc. To capture the rich textural information of the iris pattern regardless of the eye color, traditional iris recognition systems utilize near-infrared (NIR) sensors to acquire images of the iris. This, however, restricts the iris image acquisition distance to close quarters (less than 1 m). Over the last several years, there have been numerous attempts to design and implement iris recognition systems that operate at longer standoff distances ranging from 1 m to 60 m. Such long range iris acquisition and recognition systems can provide high user convenience and improved throughput. This paper reviews the state-of-the-art design and implementation of iris-recognition-at-a-distance (IAAD) systems. In this regard, the design of such a system from both the image acquisition (hardware) and image processing (algorithms) perspectives are presented. The major contributions of this paper include: (1) discussing the significance and applications of IAAD systems in the context of human recognition, (2) providing a review of existing IAAD systems, (3) presenting a complete solution to the design problem of an IAAD system, from both hardware and algorithmic perspectives, (4) discussing the use of additional ocular information, along with iris, for improving IAAD accuracy, and (5) discussing the current research challenges and providing recommendations for future research in IAAD. [ABSTRACT FROM AUTHOR]

• The establishment of a new mathematical model to realize dense light field reconstruction. • The innovative application of optical flow method. • Simple algorithm structure and relatively easy to implement. • Algorithm strong adaptability, good dense reconstruction was completed. Dense light field sampling is an important basis for refocusing, depth estimation and 3-D imaging. It is difficult to obtain high resolution dense light field with a large-scale camera array and expensive equipment. At the same time, the current storage devices and transmission bandwidth also limit this technology's post-processing and application. In order to effectively reconstruct the angle domain of the light field based on the sparse light field data, this paper analyzes the correlation and constraint relationship between the optical flow field and the motion field of multi-view images in the same scene, extends the traditional optical flow constraint equation of two-dimensional imaging to the optical flow constraint equation of four-dimensional light field, and establishes an effective mathematical model. The coordinate position of the original pixel in the new angle image is determined by coordinate search, and its intensity is reconstructed. The experimental results of multi-scene dense reconstruction show that the proposed method can reconstruct the texture, shadow and color information in the light field of a long-baseline scene with high quality. The quantitative evaluation results show that the algorithm can be applied to dense light field reconstruction of complex scenes. The algorithm in this paper is only suitable for the case of optical flow constraint in a linear light field, and the follow-up research will focus on the case of nonlinear optical flow constraint. [ABSTRACT FROM AUTHOR]

Highlights • A generalized 2D multi-level thresholding criterion function is proved rigorously by mathematical induction method. • A multi-level thresholding scheme for a RGB color image is proposed. • PSO algorithm is applied to seek to optimal threshold values in a very reasonable computational time. • The segmented image is compared with the human segmentation from BSDS300 to evaluate the experiment results quantitatively and objectively. Abstract In this paper, we propose a multi-level thresholding model based on gray-level & local-average histogram (GLLA) and Tsallis–Havrda–Charvát entropy for RGB color image. We validate the multi-level thresholding formulation by using the mathematical induction method. We apply particle swarm optimization (PSO) algorithm to obtain the optimal threshold values for each component of a RGB image. By assigning the mean values from each thresholded class, we obtain three segmented component images independently. We conduct the experiments extensively on The Berkeley Segmentation Dataset and Benchmark (BSDS300) and calculate the average four performance indices (BDE, PRI, GCE and VOI) to show the effectiveness and reasonability of the proposed method. [ABSTRACT FROM AUTHOR]

• For multi-view input, we propose the global-aware fusion to adapt the global state when fusing features from different views. • Our algorithm includes a powerful but relatively lightweight network, novel loss function, and specified training strategy. • The reconstructor exceeds existing SOTA methods while the amount of parameters is far less than the same type of algorithm. • An extra view-reduction method based on maximizing diversity for our model can be used to tradeoff cost and performance. Deep learning technology has made great progress in multi-view 3D reconstruction tasks. At present, the mainstream solutions adopt different ways to fusion the features from several views. Among them, attention-based aggregation function performs relatively well and stably, however, it still has an obvious shortcoming the strong independence of each view during predicting the weights for merging leads to a lack of adaption of the global state. In this paper, we propose a global-aware attention-based fusion approach that builds the correlation between each branch and the global feature to provide a comprehensive foundation for weights inference. On the basis of this, we design a complete reconstruction algorithm. Experiments on ShapeNet verify that our method outperforms existing SOTA methods. Furthermore, we propose a view-reduction method based on maximizing diversity and discuss the cost-performance tradeoff of our model to achieve a better performance when facing a heavy input amount and limited computational cost. [ABSTRACT FROM AUTHOR]

• A novel rank-based approach is proposed for regression tasks; • The method allows to combine predictions from a base regressor and neighborhood information; • Re-ranking strategies and weighted functions are used for better exploiting neighborhood information; • Several experiments considering various datasets, base regressors and re-ranking methods; • Significant decrease of prediction errors considering traditional and state-of-the-art regressors. Several approaches based on regression have been developed in the past few years with the goal of improving prediction results, including the use of ranking strategies. Re-ranking has been exploited and successfully employed in several applications, improving rankings by encoding the manifold structure and redefining distances among elements from a dataset. Despite the promising results observed, re-ranking has not been evaluated in regressions tasks. This paper proposes a novel, generic, and customizable framework entitled Regression by Re-ranking (RbR) , which explores the ability of re-ranking algorithms in determining relevant rankings of objects in prediction tasks. The framework relies on the integration of a base regressor, unsupervised re-ranking learning techniques, and predictions associated with nearest neighbours weighted according to their ranking positions. The RbR framework was evaluated under a rigorous experimental protocol and presented significant results in improving the prediction when compared to state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

In this paper, a robust method is presented to perform word spotting in degraded handwritten and printed document images. A new sequence matching technique, called the Flexible Sequence Matching (FSM) algorithm, is introduced for this word spotting task. The FSM algorithm was specially designed to incorporate crucial characteristics of other sequence matching algorithms (especially Dynamic Time Warping (DTW), Subsequence DTW (SSDTW), Minimal Variance Matching (MVM) and Continuous Dynamic Programming (CDP)). Along with the characteristics of multiple matching (many-to-one and one-to-many), FSM is strongly capable of skipping existing outliers or noisy elements, regardless of their positions in the target signal. More precisely, in the domain of word spotting, FSM has the ability to retrieve complete words or words that contain only a part of the query. Furthermore, due to its adaptable skipping capability, FSM is less sensitive to local variation in the spelling of words and to local degradation effects within the word image. The multiple matching capability (many-to-one, one-to-many) of FSM helps it addressing the stretching effects of query and/or target images. Moreover, FSM is designed in such a way that with little modification, its architecture can be changed into the architecture of DTW, MVM, and SSDTW and to CDP-like techniques. To illustrate these possibilities for FSM applied to specific cases of word spotting, such as incorrect word segmentation and word-level local variations, we performed experiments on historical handwritten documents and also on historical printed document images. To demonstrate the capabilities of sub-sequence matching, of noise skipping, as well as the ability to work in a multilingual paradigm with local spelling variations, we have considered properly segmented lines of historical handwritten documents in different languages and improperly as well as properly segmented words in printed and handwritten historical documents. From the comparative experimental results shown in this paper, it can be clearly seen that FSM can be equivalent or better than most DTW-based word spotting techniques in the literature while providing at the same time more meaningful correspondences between elements. [ABSTRACT FROM AUTHOR]

Ensemble pruning is an important task in supervised learning because of the performance and efficiency advantage it begets to predictive modelling. Performance based empirical comparison (primarily on accuracy) is the most common modus operandi for critical evaluation of ensembles pruned by different algorithms. Deep analysis of existing literature reveals that ensemble size is an ignored attribute while judging the quality of ensembles. In this paper, we argue that the cost-effectiveness of an ensemble is a function of both performance and size. Hence, equitable comparison of two ensembles must take into account both these attributes to judge their relative merits. Following this argument, we propose an objective function called accrual function which quantifies the difference in performance and size of two ensembles, to gauge their relative cost-effectiveness. The function can be parameterized and has nice mathematical properties. Semantic interpretations of these properties are delineated in the paper. Finally, we apply the accrual function on published results from selected publications and demonstrate its ability to beget clarity while comparing ensembles. [ABSTRACT FROM AUTHOR]

Highlights • Always considered graph edit distance (GED) is a metric if edit functions are a metric. • We discern between GED computed through edit path and graph bijection. • Triangle inequality of edit functions not necessary if GED defined by graph bijection. • Important: usually recognition ratio is maximized in non-metric edit functions. Abstract In the presentation of the graph edit distance in 1983 and other newer bibliography, authors state that it is necessary to apply the distance restrictions (non-negativity, identity of indiscernible elements, symmetry and triangle inequality) to each of the edit functions (insertion, deletion and substitution of nodes and edges) involved in the process of computing the graph edit distance to make the graph edit distance a true distance. Moreover, graph edit distance algorithms presented in the last three decades have been based on mapping the edit path that transforms a graph into the other one into a bijection of the graphs in which some null nodes have been added. In this paper, we show that the triangle inequality does not need to be imposed in each edit function if the graph edit distance is defined through an edit path; however, it is necessary if it is defined as a graph bijection. This is an important finding since the triangle inequality is the only restriction that relates different edit functions and concerns the process of tuning the edit functions given a specific application. Hence, on one hand, it would encourage research to define new algorithms based on the edit path instead of the graph bijection and, on the other hand, use edit functions without the restriction, for instance, that the sum of the costs of insertion and deletion of a pair of nodes has to be larger or equal than the cost of substituting them, which could increase the recognition ratio of a concrete application. [ABSTRACT FROM AUTHOR]

Highlights • The proposed method extends grid-based DBSCAN scalable to higher-dimensional dataset. • Cluster forest is devised to alleviate redundancies in the merging step. • HyperGrid Bitmap is used to index non-empty grids for efficient neighbor grid queries. • Experiments show performance superiority of proposed method on real/synthetic data. Abstract DBSCAN is one of clustering algorithms which can report arbitrarily-shaped clusters and noises without requiring the number of clusters as a parameter (unlike the other clustering algorithms, k -means, for example). Because the running time of DBSCAN has quadratic order of growth, i.e. O (n 2), research studies on improving its performance have been received a considerable amount of attention for decades. Grid-based DBSCAN is a well-developed algorithm whose complexity is improved to O (n log n) in 2D space, while requiring Ω(n 4/3) to solve when dimension ≥ 3. However, we find that Grid-based DBSCAN suffers from two problems: neighbour explosion and redundancies in merging, which make the algorithms infeasible in high dimensional space. In this paper we first propose a novel algorithm called GDCF which utilizes bitmap indexing to support efficient neighbour grid queries. Second, based on the concept of union-find algorithm we devise a forest-like structure, called cluster forest, to alleviate the redundancies in the merging. Moreover, we find that running the cluster forest in different orders can lead to a different number of merging operations needed to perform in the merging step. We propose to perform the merging step in a uniform random order to optimize the number of merging operations. However, for high-density database, a bottleneck could be occurred, we further propose a low-density-first order to alleviate this bottleneck. The experiments resulted on both real-world and synthetic datasets demonstrate that the proposed algorithm outperforms the state-of-the-art exact/approximate DBSCAN and suggests a good scalability. [ABSTRACT FROM AUTHOR]

Highlights • Unlike common multi-label feature selection, the proposed method derives from different cognitive standpoint. • Feature selection process is directly done on multi-label data, and there is no need to data transformation. • The proposed method tries to find label-specific features which are the most discriminative features for each label. • Also, an extension of our method is presented which selects a pre-defined number of features. • The proposed method is appropriate to both numerical and nominal features. • The proposed method is effective and fast. Abstract In multi-label data, each instance is associated with a set of labels, instead of one label. Similar to single-label data, feature selection plays an important role in improving classification performance. In multi-label classification, each class label might be specified by some particular characteristics of its own which are called label-specific features. In this paper, a fast accurate filter-based feature selection method is exclusively designed for multi-label datasets to find label-specific features. It maps the features to a multi-dimensional space based on a filter method, and selects the most salient features with the help of Pareto-dominance concepts from multi-objective optimization domain. Our proposed method can be used as online feature selection that deals with problems in which features arrive sequentially while the number of data samples is fixed. In this method, the number of features to be selected is specified during the process of feature selection. However, sometimes it is desired to predefine the number of features. For this reason, an extension of the proposed method is presented to solve this problem. To prove the performance of the proposed methods, several experiments are conducted on some multi-label datasets and the results are compared to five well-established multi-label feature selection methods. The results show the superiority of the proposed methods in terms of different multi-label classification criteria and execution time. [ABSTRACT FROM AUTHOR]

Highlights • Almost all existing attribute weighting approaches to naive Bayes are class-independent. • We propose a new class-specific attribute weighting paradigm for naive Bayes. • The resulting model is called class-specific attribute weighted naive Bayes (CAWNB). • To learn CAWNB, we propose two gradient-based learning algorithms. • The experimental results validate the effectiveness of the proposed algorithms. Abstract Due to its easiness to construct and interpret, along with its good performance, naive Bayes (NB) is widely used to address classification problems in real-world applications. In order to alleviate its conditional independence assumption, a mass of attribute weighting approaches have been proposed. However, almost all these approaches assign each attribute a same (global) weight for all classes. In this paper, we call them the general attribute weighting and argue that for NB attribute weighting should be class-specific (class-dependent). Based on this premise, we propose a new paradigm for attribute weighting called the class-specific attribute weighting, which discriminatively assigns each attribute a specific weight for each class. We call the resulting model class-specific attribute weighted naive Bayes (CAWNB). CAWNB selects class-specific attribute weights to maximize the conditional log likelihood (CLL) objective function or minimize the mean squared error (MSE) objective function, and thus two different versions are created, which we denote as CAWNBCLL and CAWNBMSE, respectively. Extensive empirical studies show that CAWNBCLL and CAWNBMSE all obtain more satisfactory experimental results compared with NB and other existing state-of-the-art general attribute weighting approaches. We believe that for NB class-specific attribute weighting could be a more fine-grained attribute weighting approach than general attribute weighting. [ABSTRACT FROM AUTHOR]

Highlights • Optimization is an important issue in machine learning because many machine learning models are reformulated as optimization problems. Different kinds of machine learning algorithms mainly focus on minimizing their emiprical loss like deep learning, logisitic regression, and support vector machine. Because data is explosively growing, it is challenging to deal with a large-scale optimization problem. We propose a fast and most condition number free algorithm for train machine learning algorithms. This algorithm has several advantages. • Contrast to previous work which approximates Hessian matrix by sketched or subsampled Hessian directly, we use the sketched or subsampled Hessian matrix as preconditioner and get the direction vector by preconditioned conjugate gradient method. Accordingly, we propose a novel stochastic second-order algorithm called Preconditioned Newton Conjugate Gradient with Sketched Hessian (PNCG). This strategy makes our algorithm almost independent on the condition number. • We numerically demonstrate the effectiveness and robustness of our algorithms in widely used machine learning algorithm Ridge Logistic Regression on several real datasets with different condition number. Our algorithms show good performance compared with existing second order methods in all cases, especially when condition number is large. Abstract Optimization is an important issue in machine learning because many machine learning models are reformulated as optimization problems. Different kinds of machine learning algorithms mainly focus on minimizing their empirical loss like deep learning, logistic regression, and support vector machine. Because data is explosively growing, it is challenging to deal with a large-scale optimization problem. Recently, stochastic second-order methods have emerged to attract much attention due to their efficiency in each iteration. These methods show good performance on training machine learning algorithms like logistic regression and support vector machine. However, the computational complexity of existing stochastic second-order methods heavily depends on the condition number of the Hessian. In this paper, we propose a new Newton-like method called Preconditioned Newton Conjugate Gradient with Sketched Hessian (PNCG). The runtime complexity of PNCG is at most logarithmic in the condition number of the Hessian. PNCG exhibits advantages over existing subsampled Newton methods especially when the Hessian matrix in question is ill-conditioned. We also show that our method has good performance on training machine learning algorithm empirically. The results show consistent improvements in computational efficiency. [ABSTRACT FROM AUTHOR]

Highlights • For Alzheimer's disease (AD) diagnosis, we consider the integration of multi-modality imaging and genetic data which encode different level of knowledge. • With a focus on taking advantage of both phenotype and genotype information, a novel structured sparsity, defined by ℓ 1, p -norm (p > 1), regularized multiple kernel learning method is designed. • An efficient block coordinate descent algorithm applicable to any case with p > 1 was derived to solve the proposed formulation. Abstract Multimodal data fusion has shown great advantages in uncovering information that could be overlooked by using single modality. In this paper, we consider the integration of high-dimensional multi-modality imaging and genetic data for Alzheimer's disease (AD) diagnosis. With a focus on taking advantage of both phenotype and genotype information, a novel structured sparsity, defined by ℓ 1, p -norm (p > 1), regularized multiple kernel learning method is designed. Specifically, to facilitate structured feature selection and fusion from heterogeneous modalities and also capture feature-wise importance, we represent each feature with a distinct kernel as a basis, followed by grouping the kernels according to modalities. Then, an optimally combined kernel presentation of multimodal features is learned in a data-driven approach. Contrary to the Group Lasso (i.e., ℓ 2, 1 -norm penalty) which performs sparse group selection, the proposed regularizer enforced on kernel weights is to sparsely select concise feature set within each homogenous group and fuse the heterogeneous feature groups by taking advantage of dense norms. We have evaluated our method using data of subjects from Alzheimer's Disease Neuroimaging Initiative (ADNI) database. The effectiveness of the method is demonstrated by the clearly improved prediction diagnosis and also the discovered brain regions and SNPs relevant to AD. [ABSTRACT FROM AUTHOR]

Highlights • We give theoretical guarantees for the block sparse representation based classifier. • A RBSRC (robust block sparse representation based classifier) framework is given. • We devise an efficient half-quadratic optimization algorithm for RBSRC. • Using RBSRC as a general platform, we propose three robust classifiers. Abstract By exploiting the low-dimensional structure of high-dimensional data, sparse representation based classifiers (SRC) has recently attracted massive attention in pattern recognition. In this paper, we study a natural generalization of SRC, i.e., block sparse representation based classifiers (BSRC), which takes into account the block structure of the dictionary. Our contributions are two-fold: (1) we provide theoretical guarantees for BSRC and theoretically show that BSRC performs perfect classification for any test sample under both cases of independent subspaces and arbitrary subspaces settings; (2) we extend BSRC and propose three robust BSRC methods based on M-estimators originating in robust statistics. This is motivated by the observation that many previous representation based classifiers utilize the mean square error (MSE) criterion as the loss function, which is sensitive to outliers and complicated noises in reality. In contrast, M-estimators has shown much stronger robustness than MSE against gross corruptions. We demonstrate the efficacy of the proposed methods through experiments on both synthetic and real-world databases for block sparse recovery, handwritten digit recognition and robust face recognition. [ABSTRACT FROM AUTHOR]

Highlights • We design a semi-supervised feature representation framework for person Re-Identification which effectively utilizes both labeled and unlabeled training data to learn a discriminative representation so that person images across disjoint camera views can be reliably matched. • A multi-view clustering method is proposed to integrate features from multiple Convolutional Neural Networks for clustering, which can give more accurate label estimation for unlabeled data. • Each of our Convolutional Neural Networks utilizes a siamese network that simultaneously computes the identification loss and verification loss, which simultaneously learns a discriminative Convolutional Neural Network embedding and a similarity metric, and thus improving pedestrian retrieval accuracy. Extensive experiments on large-scale person Re-Id datasets demonstrate the effectiveness of our method. Abstract Person Re-Identification (Re-Id) is a challenging task focusing on identifying the same person among disjoint camera views. A number of deep learning algorithms have been reported for this task in fully-supervised fashion which requires a large amount of labeled training data, while obtaining high quality labels for Re-Id is extremely time consuming. To address this problem, we propose a semi-supervised Re-Id framework by using only a small portion of labeled data and some additional unlabeled samples. This paper approaches the problem by constructing a set of heterogeneous Convolutional Neural Networks (CNNs) fine-tuned using the labeled portion, and then propagating the labels to the unlabeled portion for further fine-tuning the overall system. In this work, label estimation is a key component during the propagation process. We propose a novel multi-view clustering method, which integrates features of multiple heterogeneous CNNs to cluster and generate pseudo labels for unlabeled samples. Then we fine-tune each of the multiple heterogeneous CNNs by minimizing an identification loss and a verification loss simultaneously, using training data with both true labels and pseudo labels. The procedure is iterated until the estimation of pseudo labels no longer changes. Extensive experiments on three large-scale person Re-Id datasets demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Abstract In streaming data classification, most of the existing methods assume that all arrived evolving data are completely labeled. One challenge is that some applications where only small amount of labeled examples are available for training. Incremental semi-supervised learning algorithms have been proposed for regularizing neural networks by incorporating various side information, such as pairwise constraints or user-provided labels. However, it is hard to put them into practice, especially for non-stationary environments due to the effectiveness and parameter sensitivity of such algorithms. In this paper, we propose a novel incremental semi-supervised learning framework on streaming data. Each layer of model is comprised of a generative network, a discriminant structure and the bridge. The generative network uses dynamic feature learning based on autoencoders to learn generative features from streaming data which has been demonstrated its potential in learning latent feature representations. In addition, the discriminant structure regularizes the network construction via building pairwise similarity and dissimilarity constraints. It is also used for facilitating the parameter learning of the generative network. The network and structure are integrated into a joint learning framework and bridged by enforcing the correlation of their parameters, which balances the flexible incorporation of supervision information and numerical tractability for non-stationary environments as well as explores the intrinsic data structure. Moreover, an efficient algorithm is designed to solve the proposed optimization problem and we also give an ensemble method. Particularly, when multiple layers of model are stacked, the performance is significantly boosted. Finally, to validate the effectiveness of the proposed method, extensive experiments are conducted on synthetic and real-life datasets. The experimental results demonstrate that the performance of the proposed algorithms is superior to some state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Abstract Datasets are often collected from different resources or comprised of multiple representations (i.e., views). Multi-view clustering aims to analyze the multi-view data in an unsupervised way. Owing to the efficiency of uncovering the hidden structures of data, graph-based approaches have been investigated widely for various multi-view learning tasks. However, similarity measurement in these methods is challenging since the construction of similarity graph is impacted by several factors such as the scale of data, neighborhood size, choice of similarity metric, noise and outliers. Moreover, nonlinear relationships usually exist in real-world datasets, which have not been considered by most existing methods. In order to address these challenges, a novel model which simultaneously performs multi-view clustering task and learns similarity relationships in kernel spaces is proposed in this paper. The target optimal graph can be directly partitioned into exact c connected components if there are c clusters. Furthermore, our model can assign ideal weight for each view automatically without additional parameters as previous methods do. Since the performance is often sensitive to the input kernel matrix, the proposed model is further extended with multiple kernel learning ability. With the proposed joint model, three subtasks including construct the most accurate similarity graph, automatically allocate optimal weight for each view and find the cluster indicator matrix can be simultaneously accomplished. By this joint learning, each subtask can be mutually enhanced. Experimental results on benchmark datasets demonstrate that our model outperforms other state-of-the-art multi-view clustering algorithms. [ABSTRACT FROM AUTHOR]

Abstract Traditional collaborative representation based classification (CRC) method usually faces the challenge of data uncertainty hence results in poor performance, especially in the presence of appearance variations in pose, expression and illumination. To overcome this issue, this paper presents a CRC-based face classification method by jointly using block weighted LBP and analysis dictionary learning. To this end, we first design a block weighted LBP histogram algorithm to form a set of local histogram-based feature vectors instead of using raw images. By this means we are able to effectively decrease data redundancy and uncertainty derived from image noises and appearance variations. Second, we adopt an analysis dictionary learning model as the projection transform to construct an analysis subspace, in which a new sample is characterized with the improved sparsity of its reconstruction coefficient vector. The crucial role of the analysis dictionary learning method in CRC is revealed by its capacity of the collaborative representation in an analytic coefficient space. Extensive experimental results conducted on a set of well-known face databases demonstrate the merits of the proposed method. [ABSTRACT FROM AUTHOR]

Highlights • We promote an adaptive weighting method based on structure tensor that considers the contribution rate of each spectral band. • We propose a novel method to extract spatial details from an HSI that is more appropriate in terms of the significance of its physics and avoids spectral distortion. • We propose a deep saliency enhancement method that uses CNN to learn the feature representation from input pixels and is trained in an end-to-end manner. • We utilize the NMF algorithm to extract the spatial features provided by the enhanced saliency and spectral features provided by the original HSI and then merge these features to obtain a high-quality HSI. Abstract Limited by the existing imagery hardware and contaminated with noise or shading, spatial deterioration and spectral distortion exist in hyperspectral images (HSIs). Spectral-spatial quality enhancement was seldom addressed in a clear way albeit of first importance in HSI interpretation. In this paper, we present a promising quality enhancement method in a spectral-spatial combination framework for removing unwanted components and enhancing useful features. Our approach, called saliency detection and feature enhancement (SDFE), combines the theory of structure tensor with a deep convolutional neural network (CNN) to solve an HSI quality enhancement problem that has rarely been identified. Considering the different contribution rates of each band, an adaptive weighting method based on the eigenvalues of structure tensor is proposed to fuse the selected key band group. Then, a saliency detection method is presented to extract edge areas and corners. Owning to the success of CNN in visual-based issues, we utilize it to further enhance the saliency and obtain high-quality spatial information. To extract high-quality spectral features, the nonnegative matrix factorization (NMF) algorithm is used to extract spectral information from the original HSI. The experimental result enjoys a fact of identical materials with the similar signatures, which is useful for the subsequent application. Furthermore, our approach has a powerful influence on target detection. [ABSTRACT FROM AUTHOR]

Highlights • A geosensor data forecasting tensor framework (GDFTF) for significant societal events is proposed. • A rank increasing strategy and a sliding window strategy is used to improve the prediction accuracy. • Extensive experimental evaluations illustrate the superiority of our approach compared with the state-of-the-arts. Abstract Geosensor data forecasting has high practical value in government affairs such as prompt response and decision making. However, the spatial correlation across distinct sites and the temporal correlation within each site pose challenges to accurate forecasting. In this paper, a geosensor data forecasting tensor framework for significant societal events is proposed. Specifically, a tensor pattern is used to model the geosensor data, based on which a tensor decomposition algorithm is then developed to estimate future values of geosensor data. The proposed approach not only combines and utilizes the multi-mode correlations, but also well extracts the underlying factors in each mode of tensor and mines the multi-dimensional structures of geosensor data. In addition, a rank increasing strategy is used to determine tensor rank automatically, and a sliding window strategy is used to improve the prediction accuracy. Extensive experimental evaluations illustrate the superiority of our approach compared with the state-of-the-arts. [ABSTRACT FROM AUTHOR]

Highlights • Found and molded the relationship between the high-frequency signal of image carrier and steganalysis accuracy. • Posterior accuracy is proposed to evaluate the real accuracy of steganalysis and it can be used to improve steganalysis performance in the actual environment, because we can find a subset with a high accuracy(such as 100%) from the prediction images which have a relatively low accuracy (such as 80%). • Actually, the relationship between image carrier and steganalysis accuracy can even be used to improve steganography performance significantly. Abstract Steganalysis research is committed to distinguishing the steganography media from the normal one correctly. However, individual differences of carriers disturb the detection inevitably and greatly. Existing methods treat all detection results with the same confidence level, or prior accuracy, which may make the prior accuracy overestimate or underestimate the real result. This paper presents a novel performance evaluation method of steganalysis based on posterior accuracy. Adaptive Convolution Feature (ACF) is calculated by the adaptive convolution, then a quantitative value S based on the ACF is modeled to posterior testify and estimate the detection confidence of the image under test. By clustering of carrier noise, we classify the images which we believe they have a similar confidence of the same cluster. The distribution of ACF from each cluster indicates that the S value works well for confidence evaluation. The experimental results show that the S value can distinguish and identify high-confidence samples from the low-confidence, which will greatly improve the real performance of steganalysis. Besides, it improves the steganalysis accuracy of the whole image set by matching S values between training and prediction samples. [ABSTRACT FROM AUTHOR]

Highlights • Networks and graphs are naturally embedded in Euclidean hyperspheres. • Communicability embedding of networks/graphs reveals clusters in networks. • Nonmetric multidimensional scaling allows visualization of networks in 3D communicability space. • Communicability clusters of papers in a citation network reveal levels of mathematization. • Communicability clusters in a gene-gene network reveal genes that co-participate in cancer and other diseases. Abstract A complex network is a condensed representation of the relational topological framework of a complex system. A main reason for the existence of such networks is the transmission of items through the entities of these complex systems. Here, we consider a communicability function that accounts for the routes through which items flow on networks. Such a function induces a natural embedding of a network in a Euclidean high-dimensional sphere. We use one of the geometric parameters of this embedding, namely the angle between the position vectors of the nodes in the hyperspheres, to extract structural information from networks. First we propose a simple method for visualizing networks by reducing the dimensionality of the communicability space to 3-dimensional spheres. Secondly, we use clustering analysis to cluster the nodes of the networks based on their similarities in terms of their capacity to successfully deliver information through the network. After testing these approaches in benchmark networks and compare them with the most used clustering methods in networks we analyze two real-world examples. In the first, consisting of a citation network, we discover citation groups that reflect the level of mathematics used in their publications. In the second, we discover groups of genes that coparticipate in human diseases, reporting a few genes that coparticipate in cancer and other diseases. Both examples emphasize the potential of the current methodology for the discovery of new patterns in relational data. [ABSTRACT FROM AUTHOR]

Highlights • A Bayesian denoising hashing method is proposed for robust image retrieval. • We impose extra constraints in the variational Bayes framework to learn robust hash code. • Our method can be combined with many traditional hashing algorithms to achieve state-of-the-art performance on real-world datasets. Abstract Learning to hash is one of the most popular techniques in image retrieval, but few work investigates its robustness to noise corrupted images in which the unknown pattern of noise would heavily deteriorate the performance. To deal with this issue, we present in this paper a Bayesian denoising hashing algorithm whose output can be regarded a denoised version of the input hash code. We show that our method essentially seeks to reconstruct a new but more robust hash code by preserving the original input information while imposing extra constraints so as to correct the corrupted bits. We optimized this model in variational Bayes framework which has a closed-form update in each iteration that is more efficient than numerical optimization. Furthermore, our method can be added at the top of any original hashing layer, serving as a post-processing denoising layer with no change to previous training procedure. Experiments on three popular datasets demonstrate that the proposed method yields robust and meaningful hash code, which significantly improves the performance of state-of-the-art hash learning methods on challenging tasks such as large-scale natural image retrieval and retrieval with corrupted images. [ABSTRACT FROM AUTHOR]

Abstract In this paper, we propose the use of data symmetries, in the sense of equivalences under signal transformations, as priors for learning symmetry-adapted data representations, i.e., representations that are equivariant to these transformations. We rely on a group-theoretic definition of equivariance and provide conditions for enforcing a learned representation, for example the weights in a neural network layer or the atoms in a dictionary, to have the structure of a group and specifically the group structure in the distribution of the input. By reducing the analysis of generic group symmetries to permutation symmetries, we devise a regularization scheme for representation learning algorithm, using an unlabeled training set. The proposed regularization is aimed to be a conceptual, theoretical and computational proof of concept for symmetry-adapted representation learning, where the learned data representations are equivariant or invariant to transformations, without explicit knowledge of the underlying symmetries in the data. [ABSTRACT FROM AUTHOR]

Highlights • We proposed a weighted heterogeneous distance metric (WHDM). • We presented WHDM and Dempster–Shafer theory based k NN algorithm. • We proposed a multimodal perturbation method (RRSB) for k NN ensemble. • The effectiveness of our algorithms was shown on multiple UCI data sets and a KDD data set. Abstract In this paper, to address the issue that ensembling k -nearest neighbor (k NN) classifiers with resampling approaches cannot generate component classifiers with a large diversity, we consider ensembling k NN through a multimodal perturbation-based method. Since k NN is sensitive to the input attributes, we propose a weighted heterogeneous distance Metric (WHDM). By using a WHDM and evidence theory, a progressive k NN classifier is developed. Based on a progressive k NN, the random subspace method, attribute reduction, and Bagging, a novel algorithm termed RRSB (reduced random subspace-based Bagging) is proposed for construct ensemble classifier, which can increase the diversity of component classifiers without damaging the accuracy of the component classifiers. In detail, RRSB adopts the perturbation on the learning parameter with a weighted heterogeneous distance metric, the perturbation on the input space with random subspace and attribute reduction, the perturbation on the training data with Bagging, and the perturbation on the output target of k neighbors with evidence theory. In the experimental stage, the value of k , the different perturbations on RRSB and the ensemble size are analyzed. In addition, RRSB is compared with other multimodal perturbation-based ensemble algorithms on multiple UCI data sets and a KDD data set. The results from the experiments demonstrate the effectiveness of RRSB for k NN ensembling. [ABSTRACT FROM AUTHOR]

Due to the ability of capturing the geometry structure of data manifold, context-sensitive similarity has demonstrated impressive performances in the retrieval task. The key idea of context-sensitive similarity is that the similarity between two data points can be more reliably estimated with the local context of other points in the affinity graph. Therefore, neighborhood selection is a crucial factor for those algorithms, which affects the performance dramatically. In this paper, we propose a new algorithm called Smooth Neighborhood (SN) that mines the neighborhood structure to satisfy the manifold assumption. By doing so, nearby points on the underlying manifold are guaranteed to yield similar neighbors as much as possible. Moreover, SN is adjusted to tackle multiple affinity graphs by imposing a weight learning paradigm, and this is the primary difference compared with related works which are only applicable with one affinity graph. Finally, we integrate SN with Sparse Contextual Activation (SCA), a representative context-sensitive similarity proposed recently. Extensive experimental results and comparisons manifest that with the neighborhood structure generated by SN, the proposed framework can yield state-of-the-art performances on shape retrieval, image retrieval and 3D model retrieval. [ABSTRACT FROM AUTHOR]

The aim of this paper is to design an algorithm based on nonsmooth optimization techniques to solve the minimum sum-of-squares clustering problems in very large data sets. First, the clustering problem is formulated as a nonsmooth optimization problem. Then the limited memory bundle method [Haarala et al., 2007] is modified and combined with an incremental approach to design a new clustering algorithm. The algorithm is evaluated using real world data sets with both the large number of attributes and the large number of data points. It is also compared with some other optimization based clustering algorithms. The numerical results demonstrate the efficiency of the proposed algorithm for clustering in very large data sets. [ABSTRACT FROM AUTHOR]

Ear characteristic is a promising biometric modality that has demonstrated good biometric performance. In this paper, we investigate a novel and challenging problem to verify a subject (or user) based on the ear characteristics after undergoing ear surgery. Ear surgery is performed to reconstruct the abnormal ear structures both locally and globally to beautify the overall appearance of the ear. Ear surgery performed for both for beautification and corrections alters the original ear characteristics to the greater extent that will challenge the comparison and subsequently verification performance of the ear recognition systems. This work presents a new database of images from 211 subjects with surgically altered ear along with corresponding pre and post-surgery samples. We then propose a novel scheme for ear verification based on the features extracted using a bank of filters learnt using Topographic Locally Competitive Algorithm (T-LCA) and comparison is carried out using Robust Probabilistic Collaborative Representation Classifier (R-ProCRC). Extensive experiments are carried out on both clean (normal) and surgically altered ear database to evaluate the performance of the proposed ear verification scheme. We also present a comprehensive performance analysis by comparing the performance of the proposed ear recognition scheme with eight different state-of-the-art ear verification system. Furthermore, we also present a new scheme to detect both deformed and surgically altered ear using one-class classification. Experimental results indicate the magnitude of problem in verifying the surgically altered ears and the signifies the need for considerable research in this direction. [ABSTRACT FROM AUTHOR]

Clusters may exist in different subspaces of a multidimensional dataset. Traditional full-space clustering algorithms have difficulty in identifying these clusters. Various subspace clustering algorithms have used different subspace search strategies. They require clustering to assess whether cluster(s) exist in a subspace. In addition, all of them perform clustering by measuring similarity between points in the given feature space. As a result, the subspace selection and clustering processes are tightly coupled. In this paper, we propose a new subspace clustering framework named CSSub (Clustering by Shared Subspaces). It enables neighbouring core points to be clustered based on the number of subspaces they share. It explicitly splits candidate subspace selection and clustering into two separate processes, enabling different types of cluster definitions to be employed easily. Through extensive experiments on synthetic and real-world datasets, we demonstrate that CSSub discovers non-redundant subspace clusters with arbitrary shapes in noisy data; and it significantly outperforms existing state-of-the-art subspace clustering algorithms. [ABSTRACT FROM AUTHOR]

In this paper, we investigate the non-zero representation coefficients of Locally Linear KNN (LLKNN) and propose a nonnegative extension of LLKNN (NLLKNN) model for image recognition, where representation coefficients are restricted to be nonnegative to avoid meaningless and unreasonable negative coefficients. A multiplicative iterative algorithm with proof of convergence is proposed to solve the proposed NLLKNN model. Then NLLKNN based classifier (NLLKNNC) and Nonnegative Locally Linear Nearest Mean Classifier (NLLNMC) are proposed. We also investigate the robustness of NLLKNNC and NLLNMC to noises and occlusions. The effectiveness of the proposed methods is evaluated on several image recognition tasks such as scene recognition and face recognition. Extensive experimental results show that the proposed algorithm converges very fast and the proposed methods outperform some representative image recognition methods. [ABSTRACT FROM AUTHOR]

In this work we describe a fast and stable algorithm for the computation of the orthogonal moments of an image. Indeed, orthogonal moments are characterized by a high discriminative power, but some of their possible formulations are characterized by a large computational complexity, which limits their real-time application. This paper describes in detail an approach based on recurrence relations, and proposes an optimized Matlab implementation of the corresponding computational procedure, aiming to solve the above limitations and put at the community’s disposal an efficient and easy to use software. In our experiments we evaluate the effectiveness of the recurrence formulation, as well as its performance for the reconstruction task, in comparison to the closed form representation, often used in the literature. The results show a sensible reduction in the computational complexity, together with a greater accuracy in reconstruction. In order to assess and compare the accuracy of the computed moments in texture analysis, we perform classification experiments on six well-known databases of texture images. Again, the recurrence formulation performs better in classification than the closed form representation. More importantly, if computed from the GLCM of the image using the proposed stable procedure, the orthogonal moments outperform in some situations some of the most diffused state-of-the-art descriptors for texture classification. [ABSTRACT FROM AUTHOR]

Clustering is an important technique to deal with large scale data which are explosively created in internet. Most data are high-dimensional with a lot of noise, which brings great challenges to retrieval, classification and understanding. No current existing approach is “optimal” for large scale data. For example, DBSCAN requires O ( n 2 ) time, Fast-DBSCAN only works well in 2 dimensions, and ρ -Approximate DBSCAN runs in O ( n ) expected time which needs dimension D to be a relative small constant for the linear running time to hold. However, we prove theoretically and experimentally that ρ -Approximate DBSCAN degenerates to an O ( n 2 ) algorithm in very high dimension such that 2 D  >  >  n . In this paper, we propose a novel local neighborhood searching technique, and apply it to improve DBSCAN, named as NQ-DBSCAN, such that a large number of unnecessary distance computations can be effectively reduced. Theoretical analysis and experimental results show that NQ-DBSCAN averagely runs in O ( n * log ( n )) with the help of indexing technique, and the best case is O ( n ) if proper parameters are used, which makes it suitable for many realtime data. [ABSTRACT FROM AUTHOR]