Search

Mixture faults detection is meaningful for gasoline engines because proper mixture is the basic prerequisite for healthy running of a combustion engine. Among existing methods for faults detection, the data-driven trend analysis technique is widely used due to the simplicity and efficiency in time-domain. The CUSUM (Cumulative Sum Of Errors) algorithm is good at real-time trend extraction, but it's easy to be costly on the fuel trim signal during the engine in normal working conditions, which will increase battery energy consumption because engine failure is rarely occurs. Hence, the conventional treatment methods of artifacts in the CUSUM algorithm are modified by means of decay function and detection time analysis. The thresholds are tuned according to the characteristics of artifacts instead of residual variability, which leads to better results of trend extraction and less computation. Then, the revised CUSUM algorithm is used for monitoring the mixture abnormal behaviors, and the mixture faults can be detected in real time through analyzing the variation features of fuel trim signal. The lightweight faults detector using the advanced CUSUM algorithm (FD-A-CUSUM) is evaluated on the experimental data collected from a Ford engine. The validation results show that while engine works under normal conditions, the computation of FD-A-CUSUM has decreased by 72.79 % in comparison with the detection method using the original CUSUM algorithm (FD-O-CUSUM), and the false alarm ratio of FD-A-CUSUM is 3.37 %. Futhermore, the detection results of FD-A-CUSUM for two leakage faults have achieved 91.18 % test accuracy. [ABSTRACT FROM AUTHOR]

Low-rank representation (LRR) has been successfully applied to subspace clustering. However, the nuclear norm in the standard LRR is not optimal for approximating the rank function in many real-world applications. Meanwhile, the L21 norm in LRR also fails to characterize various noises properly. To address the above issues, we propose an improved LRR method, which achieves low rank property via the new formulation with weighted Schatten-p norm and Lq norm (WSPQ). Specifically, the nuclear norm is generalized to be the Schatten-p norm and different weights are assigned to the singular values, and thus it can approximate the rank function more accurately. In addition, Lq norm is further incorporated into WSPQ to model different noises and improve the robustness. An efficient algorithm based on the inexact augmented Lagrange multiplier method is designed for the formulated problem. Extensive experiments on face clustering and motion segmentation clearly demonstrate the superiority of the proposed WSPQ over several state-of-the-art methods. [ABSTRACT FROM AUTHOR]

In order to improve the estimate performance of distributed sensor networks, we introduce an improved consensus-based distributed filtering algorithm. Firstly, sensors obtain the local estimates using their own measurements. Then through utilizing the data of neighbor nodes to update the local estimates, estimates in the networks can reach dynamic average consensus. Based on studying the value of consensus step size, we give a sufficient condition for the convergence of the algorithm and discusses the influences of the consensus step size and the detection probability on the accuracy and consensus of estimation. The numerical simulation demonstrates that the algorithm proposed in this paper can improve the accuracy and consensus of estimation, and it is more robust with incomplete measurements. [ABSTRACT FROM AUTHOR]

Highlights: [•] Class label information is considered in the procedure of feature extraction. [•] Local structure is kept in the feature space. [•] Locality preserving may produce high between-class overlaps by using the k nearest neighbor criterion. [•] Maximizing the dissimilarity of samples in a manifold is beneficial for classification. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, we propose a local descriptor, called Weber Local Binary Pattern (WLBP 1 [1] WLBP is the abbreviation of Weber Local Binary Pattern. ), which effectively combines the advantages of WLD and LBP. Specifically, WLBP consists of two components: differential excitation and LBP. The differential excitation extracts perception features by Weber's law, while the LBP (Local Binary Pattern) can describe local features splendidly. By computing the two components, we obtain wo images: differential excitation image and LBP image, from which a WLBP histogram is constructed. The differential excitation was extended by bringing in Laplacian of Gaussian (LoG), which makes WLBP robust to noise. By designing a new quantization method, the discriminabilty of WLBP was enhanced. The proposed method is evaluated on the face recognition problem under different challenges. Experimental results show that WLBP performs better than WLD and LBP. Meanwhile, it is robust to time, facial expressions, lightings, pose and noise. We also conduct experiments on Brodatz and KTH-TIPS2-a texture databases, which demonstrate that WLBP is a powerful texture descriptor. [Copyright &y& Elsevier]

Single-sample face recognition is a very challenging problem, where each person has only one labeled training sample. It is difficult to describe unknown facial variations. In this paper, we propose a shared generative adversarial network (SharedGAN) to expand the gallery dataset. Benefiting from the shared decoding network, SharedGAN requires only a small number of training samples. After obtaining the generated samples, we join them into a large public dataset. Then, a deep convolutional neural network is trained on the new dataset. We use the well-trained model for feature extraction. With the deep convolutional features, a simple softmax classifier is trained. Our method has been evaluated on AR, CMU-PIE, and FERET datasets. Experimental results demonstrate the effectiveness of SharedGAN and show its robustness for single sample face recognition. [ABSTRACT FROM AUTHOR]

Abstract: In this work, we propose a linear representation based face recognition (FR) method incorporating locality information from both spatial features and training samples. Instead of holistic face images, the proposed method is conducted on the spatial pyramid local patches, which are aggregated by a Bayesian based fusion method. The locality constraint on the representation coefficients leads to an approximately sparse representation, which effectively explores the discriminative nature of spatial local features. Different from the sparse representation based classification (SRC) exposing an -norm constraint on the coefficients, the proposed locality constrained representation based classification (LCRC) is formulated with a computationally efficient -norm. The proposed method is robust to two crucial problems in face recognition: occlusion and lack of training data. A simple locality based concentration index (LCI) is defined to measure the reliability of each local patch, by which not only the heavily corrupted patches but also the less discriminant ones are rejected. Due to the use of both local patches and the locality constraint, less training data are required by the proposed method. Based on the locality constrained representation, we present three algorithms which outperform the state-of-the-art on the AR and Extended Yale B datasets for both the occlusion and single sample per person (SSPP) problems. [Copyright &y& Elsevier]

Abstract: Kernel discriminant analysis (KDA) is an effective statistical method for dimensionality reduction and feature extraction. However, traditional KDA methods suffer from the small sample size problem. Moreover, they endure the Fisher criterion that is nonoptimal with respect to classification rate. This paper presents a variant of KDA that deals with both of the shortcomings in an efficient and cost effective manner. The key to the approach is to use simultaneous diagonalization technique for optimization and meanwhile utilize a modified Fisher criterion that it is more closely related to classification error. Extensive experiments on face recognition task show that the proposed method is an effective nonlinear feature extractor. [Copyright &y& Elsevier]

Abstract: Tracking infrared pedestrian targets is a crucial part in video surveillance. Many factors make this problem decidedly non-linear and non-Gaussian, and the appropriate solution at present is based on the particle filter technique which is powerful and simple to implement. But in many cases, the traditional particle filter tracking algorithm fails to track the targets robustly and accurately. To solve these problems, a modified particle filter algorithm that combines intensity and edge cues is proposed. The algorithm firstly extracts the intensity cue and edge cue of the target based on the visual models which are originally learnt from the first frame and will be updated during the tracking process according to an automatic model updating strategy. Secondly, these two cues are combined into the particle filter framework by an adaptive integration scheme. Furthermore, its performance is evaluated with real-world infrared pedestrian sequences and extensive experimental results show that the presented method can track the infrared pedestrian more effectively and reliably than the traditional particle filter algorithm. [Copyright &y& Elsevier]

• Network is dedicated to learn complementary feature representations. • Multi-task learning with classification and pair-wise ranking. • Network achieves competitive results compared to state-of-the-arts. Person re-identification aims at establishing the identity of a pedestrian from a gallery that contains images of people obtained from a multi-camera system, which has many applications in video surveillance for public security and safety. Many challenges such as occlusions, drastic lighting and pose variations across the camera views, and noise make this task highly challenging. While most approaches focus on learning features and metrics to derive better representations, we hypothesize that both local and global contextual cues are crucial for an accurate identity matching. To this end, we propose a Feature Mask Network (FMN) that takes advantage of ResNet high-level features to predict a feature map mask and then imposes it on the low-level features to dynamically re-weight different object parts for a complementary feature representation. This serves as an attention mechanism by allowing the network to focus on local details selectively. We frame the network training as a multi-task objective optimization, which further improves the learned feature descriptions. We conduct experiments on Market-1501, DukeMTMC-reID and CUHK03 datasets, where the proposed approach respectively achieves significant improvements and competitive results when compared to the state-of-the-art. [ABSTRACT FROM AUTHOR]

Hyperspectral unmixing is one of the most important techniques in the remote sensing image analysis tasks. In recent decades, nonnegative matrix factorization (NMF) has been shown to be effective for hyperspectral unmixing due to the strong discovery of the latent structure. Most NMFs put emphasize on the spectral information, but ignore the spatial information, which is very crucial for analyzing hyperspectral data. In this paper, we propose an improved NMF method, namely NMF with region sparsity learning (RSLNMF), to simultaneously consider both spectral and spatial information. RSLNMF defines a new sparsity learning model based on a small homogeneous region that is obtained via the graph cut algorithm. Thus RSLNMF is able to explore the relationship of spatial neighbor pixels within each region. An efficient optimization scheme is developed for the proposed RSLNMF, and its convergence is theoretically guaranteed. Experiments on both synthetic and real hyperspectral data validate the superiority of the proposed method over several state-of-the-art unmixing approaches. [ABSTRACT FROM AUTHOR]

The horizontal axis-symmetrical nature of faces is useful and interesting, which has successfully applied in face detection. In general, images of faces are not strictly captured under a frontal and natural pose. It implies that extra training samples could be generated by means of the symmetry of the face. In this paper, we develop a framework that fuses virtual mirror synthesized training samples as bases and hierarchical multi-scale local binary patterns (LBP) features for classification. More specifically, in the first stage of proposed method, sampling uncertainty of the linear approximation model is alleviated effectively by constructing extra synthesized mirror training samples which generated from original images. Subsequently, in the second stage, features are extracted from the above dictionary using a hierarchical multi-scale LBP scheme. In the third stage, we combine the synthesized samples and original ones together to describe a test sample and simultaneously determine the relatively informative training samples by means of exploiting the reconstruction deviation of all the dictionary atoms. Besides, the introduced sparse coding process with weak l constraint has superior competitiveness that the accuracy has been improved, while the complexity has been fallen. Ultimately, the following step determines again a new decomposition coefficient of remaining samples, which makes final decision of the classification. Experimental results that are engaged in various benchmark face databases have demonstrated the effectiveness of our algorithm. [ABSTRACT FROM AUTHOR]

As is well known, most of the sparse representation based classification (SRC) schemes are not the standard form of SR model, that is, the dictionary in SRC model is not a strict one compare to sparse representation model. Actually, a strict and standard dictionary will create sparser coefficient, but it might not achieve better classification accuracy in SRC model. Therefore, constructing a rational and optimal dictionary for SRC model is a challenging task for us. Understanding the good performance of such unconventional dictionaries demands new algorithmic and analytical techniques. In this paper, we propose a novel SRC fusion method using hierarchical multi-scale local binary patterns (LBP) and greedy search strategy (SRC-GSLBP) for face recognition. The proposed method involves three aspects: dictionary optimization, training variables selection and classification strategy, sparse coding coefficient decomposing. First of all, we get an optimized dictionary through extracting hierarchical multi-scale LBP features from the original training samples. Second, the training variables selection and classification strategy aim to represent a query sample as a linear combination of the most informative training samples, and exploits an optimal representation of training samples from the classes with major relevant contributions. Instead of eliminating several classes at one time, we choose eliminating classes one by one with greedy search (GS) sparse coding process until the predefined termination condition is satisfied. The final remaining training samples are used to produce a best representation of the test sample and to perform classification. In the context of the proposed method, an important goal is to select a subset of variables that accomplishes one objective: the provision of a descriptive representation for sparse category knowledge structure. We develop a heuristic learning strategy to achieve this goal. Experimental results conducted on the ORL, FERET and AR face databases demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Person Re-Identification (ReID) has witnessed tremendous improvements with the help of deep convolutional neural networks (CNN). Nevertheless, because different fields have their characteristics, most existing methods encounter the problem of poor generalization ability to invisible people. To address this problem, based on the relationship between the temporal and camera position, we propose a robust and effective training strategy named temporal smoothing dynamic re-weighting and cross-camera learning (TSDRC). It uses robust and effective algorithms to transfer valuable knowledge of existing labeled source domains to unlabeled target domains. In the target domain training stage, TSDRC iteratively clusters the samples into several centers and dynamically re-weights unlabeled samples from each center with a temporal smoothing score. Then, cross-camera triplet loss is proposed to fine-tune the source domain model. Particularly, to improve the discernibility of CNN models in the source domain, generally shared person attributes and margin-based softmax loss are adapted to train the source model. In terms of the unlabeled target domain, the samples are clustered into several centers iteratively and the unlabeled samples are dynamically re-weighted from each center. Then, cross-camera triplet loss is proposed to fine-tune the source domain model. Comprehensive experiments on the Market-1501 and DukeMTMC-reID datasets demonstrate that the proposed method vastly improves the performance of unsupervised domain adaptability. [ABSTRACT FROM AUTHOR]

Most existing action recognition approaches directly leverage the video-level features to recognize human actions from videos. Although these methods have made remarkable progress, the accuracy is still unsatisfied. When the test video involves complex backgrounds and activities, existing methods usually suffer from a significant drop in accuracy. Human action is inherently a high-level concept. Merely applying a video classification model without a detailed semantic understanding of the video content, e.g., objects, scene context, object motions, object interactions, is inadequate to tackle the challenges for action recognition. Fine-level semantic understanding of videos generates elementary semantic concepts from the raw video data, such as the semantics of objects and background regions. It can be employed to bridge the gap between the raw video data and the high-level concept of human actions. In this work, we leverage dense semantic segmentation masks, which encode rich semantic details, provide extra information for the network training, and improve the performance of action recognition. We propose a novel deep architecture which is named as Dense Semantics-Assisted Convolutional Neural Networks (DSA-CNNs) to effectively utilize dense semantic information of video by a bottom-up attention way in the spatial stream, while by the way of branch fusion in the temporal stream. To verify the effectiveness of our approach, we conduct extensive experiments on publicly available datasets – UCF101, HMDB51, and Kinetics. The experimental results demonstrate that our approach substantially improves existing methods and achieves very competitive performance. It also shows that our approach is superior to other related methods that utilize extra information for action recognition. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, an efficient feature extraction method, named local structure preserving discriminant analysis (LSPDA), is presented. LSPDA constructs the local scatter and the between-class scatter to characterize the sub- and multi-manifold information respectively. More specifically, the local structure is constructed according to a certain kind of similarity between data points which takes special consideration of both the local information and the class information based on a parameter-free neighborhood decision rule, and the between-class structure is constructed according to the importance degrees of the not-same-class points measured by a strictly monotonically decreasing function. After the local scatter and the between-class scatter have been characterized, the novel feature extraction criterion is derived via maximizing the difference between the between-class scatter and the local scatter. Experimental results on the Wine dataset, AR, FERET, CMU PIE, ORL and LFW face databases show the effectiveness of the proposed method. [Copyright &y& Elsevier]

Abstract: In this paper, a novel supervised dimensionality reduction algorithm called discriminant similarity and variance preserving projection (DSVPP) is presented for feature extraction and recognition. More specifically, we redefine the intrinsic graph and penalty graph to model the intra-class compactness and inter-class separability of data points, where the intrinsic graph characterizes the similarity information of the same-class points and the penalty graph characterizes the variance information of the not-same-class points. Using the two graphs, the within-class scatter and the between-class scatter are computed, and then a concise feature extraction criterion is raised via minimizing the difference between them. Experimental results on the Wine data set, ORL, FERET and AR face databases show the effectiveness of the proposed method. [Copyright &y& Elsevier]

Low-rank tensor recovery faces challenges in accurately defining the low-rankness of a tensor. Most existing definitions typically focus on one domain alone — either the original or frequency domain. Additionally, certain definitions often exhibit limitations in their sensitivity to orientation variation. To overcome these challenges, we define a novel tensor rank, the Orientation Invariant Hybrid Rank (OIHR). This rank fuses rank information across all orientations in both frequency and original domains. Employing its convex approximation, the Orientation Invariant Hybrid Nuclear Norm (OIHNN), we propose a general tensor recovery model. We further explore the statistical performance of the estimator based on this model, establishing a deterministic upper bound on the estimation error under generic noise. Furthermore, non-asymptotic upper bounds under Gaussian noise are separately derived for two specific cases: tensor compressive sensing and tensor completion. Finally, we propose the algorithm to solve the model. Extensive experiments on both synthetic and real data are conducted to validate the statistical guarantees and verify the effectiveness of our algorithm. • Our tensor rank fuses low-rankness on all orientations in frequency–original domain. • Tensor recovery algorithm is obtained and outperforms existing convex methods. • Statistical Guarantees are established for estimated error. [ABSTRACT FROM AUTHOR]

• A novel Skeleton Gait Energy Image (SGEI) is proposed for gait recognition. • SGEIs are robust in the unconstrained environments. • A two-branch multi-stage CNN is applied for skeleton extraction. • The Singular Value Decomposition (SVD) is applied on SGEIs for view normalization. • Hybriding GEIs and SGEIs can enhance the robustness of gait recognition. Gait features have been widely applied in human identification. The commonly-used representations for gait recognition can be roughly classified into two categories: model-free features and model-based features. However, due to the view variances and clothes changes, model-free features are sensitive to the appearance changes. For model-based features, there is great difficulty in extracting the underlying models from gait sequences. Based on the confidence maps and the part affinity fields produced by a two-branch multi-stage CNN network, a new model-based representation, Skeleton Gait Energy Image (SGEI), has been proposed in this paper. Another contribution is that a hybrid representation has been produced, which uses SGEI to remedy the deficiency of model-free features, Gait Energy Image (GEI) for instance. The experimental performances indicate that our proposed methods are more robust to the cloth changes, and contribute to increasing the robustness of gait recognition in the unconstrained environments with view variances and clothes changes. [ABSTRACT FROM AUTHOR]

Highlights • Our proposed approach eliminates the dependence on manually labeled knowledge. • We propose to extract the semantic refinement subcategories from untagged corpora has a better scalability. • We propose a new unified objective function to jointly learn the classifiers for categories and subcategories. • Our approach shows substantial improvement over existing weakly supervised state-of-the-art baseline methods. Abstract Studies show that refining real-world categories into semantic subcategories contributes to better image modeling and classification. Previous image sub-categorization work relying on labeled images and WordNet's hierarchy is labor-intensive. To tackle this problem, in this work, we extract textual and visual features to automatically select and subsequently classify web images into semantic rich categories. The following two major challenges are well studied: (1) noise in the labels of subcategories derived from the general corpus; (2) noise in the labels of images retrieved from the web. Specifically, we first obtain the semantic refinement subcategories from the text perspective and remove the noise by using the relevance-based approach. To suppress the search error induced noisy images, we then formulate image selection and classifier learning as a multi-instance learning problem and propose to solve the employed problem by the cutting-plane algorithm. The experiments show significant performance gains by using the generated data of our approach on image categorization tasks. The proposed approach also consistently outperforms existing weakly supervised and web-supervised approaches. [ABSTRACT FROM AUTHOR]

• Explore deep metric learning in the presence of label noise. • Make a trade-off between informativeness and cleanliness by mining semi-hard samples. • Apply a re-ranking strategy to label correction based on the feature space. • Achieve state-of-the-art performance on three synthetic and three real-world noisy benchmarks. Due to the memorization effect of Deep Neural Networks, training with noisy labels typically results in performance degradation. Existing methods have devoted numerous efforts to training a robust model. However, rare literature pays attention to deep metric learning in the presence of label noise due to the challenge of discriminating between noisy examples and informative (hard) ones. To this end, we propose a simple yet effective approach Co-mining to build a robust and discriminative feature space. Specifically, more confident yet fewer instances are selected to guarantee the label quality and make up a relatively reliable gallery for exploring under-exploited patterns. Considering that some unexpected noisy instances are inevitably registered as clean ones as training progresses, we make a trade-off between cleanliness and informativeness. A triplet loss, pulling positive pairs closer and pushing negative pairs apart, is imposed on semi-hard samples rather than the hardest counterparts. Lastly, our method employs a threshold to distinguish out-of-distribution noisy images and discards them directly. We also use a re-ranking strategy for in-distribution noisy data to correct corrupted labels based on their consistently high predicted probabilities. Comprehensive experiments on various synthetic and real-world benchmarks demonstrate the superiority of our Co-mining. Code is available at https://github.com/NUST-Machine-Intelligence-Laboratory/Co-mining. [ABSTRACT FROM AUTHOR]

The goal of this work is to automatically collect a large number of highly relevant natural images from Internet for given queries. A novel automatic image dataset construction framework is proposed by employing multiple query expansions. In specific, the given queries are first expanded by searching in the Google Books Ngrams Corpora to obtain a richer semantic descriptions, from which the visually non-salient and less relevant expansions are then filtered. After retrieving images from the Internet with filtered expansions, we further filter noisy images by clustering and progressively Convolutional Neural Networks (CNN) based methods. To evaluate the performance of our proposed method for image dataset construction, we build an image dataset with 10 categories. We then run object detections on our image dataset with three other image datasets which were constructed by weak supervised, web supervised and full supervised learning, the experimental results indicated the effectiveness of our method is superior to weak supervised and web supervised state-of-the-art methods. In addition, we do a cross-dataset classification to evaluate the performance of our dataset with two publically available manual labelled dataset STL-10 and CIFAR-10. [ABSTRACT FROM AUTHOR]

• Mitigate the harmful effects of label noise and capture discriminative feature representations simultaneously • Decouple the joint optimization to promote the generalization performance • Refurbish training procedure with accumulated predictions and consistency • Achieves state-of-the-art performance for the web-supervised fine-grained recognition task Due to DNNs' memorization effect, label noise lessens the performance of the web-supervised fine-grained visual categorization task. Previous literature primarily relies on small-loss instances for subsequent training. The current state-of-the-art approach JoCoR additionally employs explicit consistency constraints to make clean samples more confident. However, a joint loss designed for both sample selection criteria and parameter updating is not competent for training a robust model in the presence of web noise. Especially, false positives are assigned with larger weights, causing the model to pay more attention to misclassified noisy images. Besides, leveraging weight decay to forget discarded noisy instances is too slow and implicit to take effect. Therefore, we propose a simple yet effective approach named MS-DeJOR (M ulti- S cale training with De coupled J oint O ptimization and R efurbishment). In contrast to JoCoR, we decouple sample selection from training procedure to handle the above problems. Specifically, a negative entropy term is applied to prevent false positives from being overemphasized. The model can explicitly forget those samples identified as noise by imposing such a regularization term on all training data. Furthermore, we use accumulated predictions to refurbish the noisy labels and re-weight training images to boost the model performance. A multi-scale feature enhancement module is adopted to extract discriminative and subtle feature representations. Extensive experiments show that MS-DeJOR yields state-of-the-art performances on three web-supervised fine-grained datasets, demonstrating the effectiveness of our approach. The data and source code have been available at https://github.com/msdejor/MS-DeJOR. [ABSTRACT FROM AUTHOR]

In this paper, we present a fast non-blind deconvolution method for restoring blurred images contaminated by Poisson noise. The problem is formulated by finding the minimizer of the negative logarithmic Poisson log-likelihood combined with the total variation (TV). To attack the challenging task, we adopt the well-known variable splitting and penalty technique to convert the problem into two easier sub-problems: one is a modified TV regularized deconvolution and the other is a simple convex optimization problem. Experimental results show that the proposed method runs very fast and the quality of the restored image is comparable with that of some state of the art methods. [ABSTRACT FROM AUTHOR]

This article presents a simple yet effective face recognition method, called local structure-based sparse representation classification (LS_SRC). Motivated by the “divide-and-conquer” strategy, we first divide the face into local blocks and classify each local block, then integrate all the classification results to make the final decision. To classify each local block, we further divide each block into several overlapped local patches and assume that these local patches lie in a linear subspace. This subspace assumption reflects the local structure relationship of the overlapped patches, making sparse representation-based classification (SRC) feasible even when encountering the single-sample-per-person (SSPP) problem. To lighten the computing burden of LS_SRC, we further propose the local structure-based collaborative representation classification (LS_CRC). Moreover, the performance of LS_SRC and LS_CRC can be further improved by using the confusion matrix of the classifier. Experimental results on four public face databases show that our methods not only generalize well to SSPP problem but also have strong robustness to occlusion; little pose variation; and the variations of expression, illumination, and time. [ABSTRACT FROM AUTHOR]

Abstract: Recently, Universum data that does not belong to any class of the training data, has been applied for training better classifiers. In this paper, we address a novel boosting algorithm called AdaBoost that can improve the classification performance of AdaBoost with Universum data. AdaBoost chooses a function by minimizing the loss for labeled data and Universum data. The cost function is minimized by a greedy, stagewise, functional gradient procedure. Each training stage of AdaBoost is fast and efficient. The standard AdaBoost weights labeled samples during training iterations while AdaBoost gives an explicit weighting scheme for Universum samples as well. In addition, this paper describes the practical conditions for the effectiveness of Universum learning. These conditions are based on the analysis of the distribution of ensemble predictions over training samples. Experiments on handwritten digits classification and gender classification problems are presented. As exhibited by our experimental results, the proposed method can obtain superior performances over the standard AdaBoost by selecting proper Universum data. [Copyright &y& Elsevier]

Boosting algorithms, especially AdaBoost, have attracted great attention in computer vision. In the early version of boosting algorithms, the weak classifier selection and the strong classifier learning are linked together. It has been demonstrated that decoupling of these two processes can provide more flexibility for training a better classifier. In these studies, linear discriminant analysis (LDA) has been adopted to select weak classifiers independently based on class separability rather than a training error that occurs normally in AdaBoost. It is observed that LDA is successful only if a large number of labeled training samples is available. However, a large-scale labeled training set is not always available in many computer vision applications such as object classification. To tackle this problem, this paper proposes semisupervised subspace learning combined with a boosting framework for object classification, through which unlabeled data can participate in the boosting training to compensate for the lack of enough labeled data. With the proposed framework, this paper develops three various approaches that utilize unlabeled data in different ways. According to the experiments on several public image data sets, the proposed methods achieve superior performance over AdaBoost and existing semisupervised algorithms. [ABSTRACT FROM PUBLISHER]

Abstract: This paper is interested in the problem of asymptotic stability analysis of Bidirectional Associative Memory (BAM) neural networks with time-varying delays via delta operator approach. The delays are assumed to exist in the nonlinear synaptic connection between different neural fields. Based on the Lyapunov–Krasovskii functional in delta domain, a new delay-dependent criterion for analyzing the asymptotic stability of BAM neural networks is obtained. Some previous results of continuous and discrete BAM systems are unified into the delta operator system framework due to the favorable numerical properties and the quasi-continuous performance of delta operator approach at high sampling rates. Since the sampling period is an explicit parameter in the results, it can be regulated to analyze the stability of systems. Numerical examples are presented to demonstrate the effectiveness of the developed theoretical results. [Copyright &y& Elsevier]

The least trimmed sum of squares (LTS) regression estimation criterion is a robust statistical method for model fitting in the presence of outliers. Compared with the classical least squares estimator, which uses the entire data set for regression and is consequently sensitive to outliers, LTS identifies the outliers and fits to the remaining data points for improved accuracy. Exactly solving an LTS problem is NP-hard, but as we show here, LTS can be formulated as a concave minimization problem. Since it is usually tractable to globally solve a convex minimization or concave maximization problem in polynomial time, inspired by refid="ref1"/, we instead solve LTS' approximate complementary problem, which is convex minimization. We show that this complementary problem can be efficiently solved as a second order cone program. We thus propose an iterative procedure to approximately solve the original LTS problem. Our extensive experiments demonstrate that the proposed method is robust, efficient and scalable in dealing with problems where data are contaminated with outliers. We show several applications of our method in image analysis. [ABSTRACT FROM AUTHOR]

• Adapt to dynamic changes in search results and carry out visual disambiguation. • A saliency-guided deep MIL network to remove outliers and learn visual disambiguation. • Identify appropriate visual senses for sense-specific image collection. • Be used as a pre-step before learning from the web for improving the performance. Due to the high cost of manual annotation, learning directly from the web has attracted broad attention. One issue that limits the performance of current webly supervised models is the problem of visual polysemy. In this work, we present a novel framework that resolves visual polysemy by dynamically matching candidate text queries with retrieved images. Specifically, our proposed framework includes three major steps: we first discover and then dynamically select the text queries according to the keyword-based image search results, we employ the proposed saliency-guided deep multi-instance learning (MIL) network to remove outliers and learn classification models for visual disambiguation. Compared to existing methods, our proposed approach can figure out the right visual senses, adapt to dynamic changes in the search results, remove outliers, and jointly learn the classification models. Extensive experiments and ablation studies on CMU-Poly-30 and MIT-ISD datasets demonstrate the effectiveness of our proposed approach. [ABSTRACT FROM AUTHOR]

Barrier coverage is attractive for many practical applications of directional sensor networks. Power conservation is one of the important issues in directional sensor networks. In this article, we address energy-efficient barrier coverage for directional sensor networks with mobile sensors. First, we derive the critical condition for mobile deployment. We assume that a number of stationary directional sensors are placed independently and randomly following a Poisson point process in a two-dimensional rectangular area. Our analysis shows that the critical condition only depends on the deployment density ( λ ) and the sensing radius (r). When the initial deployment satisfies λ < 8 ln 2 / r ^ 2 , barrier gaps may exist, so we need to redeploy mobile sensors to improve the barrier coverage. Then, we propose an energy-efficient barrier repair algorithm to construct an energy-efficient barrier to detect intruders moving along restricted crossing paths in the target area. Through extensive simulations, the results show that the energy-efficient barrier repair algorithm improves the barrier coverage and prolongs the network lifetime by minimizing the maximum sensor moving distance. And in comparison with the energy-efficient barrier coverage algorithm (previous works), the energy-efficient barrier repair algorithm increases by 18% of network lifetime on average. [ABSTRACT FROM AUTHOR]