Sparsity Concentration Index Research Articles

Dynamic Sparse Subspace Clustering for Evolving High-Dimensional Data Streams.

In an era of ubiquitous large-scale evolving data streams, data stream clustering (DSC) has received lots of attention because the scale of the data streams far exceeds the ability of expert human analysts. It has been observed that high-dimensional data are usually distributed in a union of low-dimensional subspaces. In this article, we propose a novel sparse representation-based DSC algorithm, called evolutionary dynamic sparse subspace clustering (EDSSC). It can cope with the time-varying nature of subspaces underlying the evolving data streams, such as subspace emergence, disappearance, and recurrence. The proposed EDSSC consists of two phases: 1) static learning and 2) online clustering. During the first phase, a data structure for storing the statistic summary of data streams, called EDSSC summary, is proposed which can better address the dilemma between the two conflicting goals: 1) saving more points for accuracy of subspace clustering (SC) and 2) discarding more points for the efficiency of DSC. By further proposing an algorithm to estimate the subspace number, the proposed EDSSC does not need to know the number of subspaces. In the second phase, a more suitable index, called the average sparsity concentration index (ASCI), is proposed, which dramatically promotes the clustering accuracy compared to the conventionally utilized SCI index. In addition, the subspace evolution detection model based on the Page-Hinkley test is proposed where the appearing, disappearing, and recurring subspaces can be detected and adapted. Extinct experiments on real-world data streams show that the EDSSC outperforms the state-of-the-art online SC approaches.

Local adaptive joint sparse representation for hyperspectral image classification

In this paper, a local adaptive joint sparse representation (LAJSR) model is proposed for the classification of hyperspectral remote sensing images. It improves the original joint sparse representation (JSR) method in both the signal and dictionary construction phase and sparse representation phase. Given a testing pixel, a similar signal set is constructed by picking a few of the most similar pixels from its spatial neighborhood. The original training dictionary consists of training samples from different classes and is extended by adding spatial neighbors of each training sample. A local adaptive dictionary is built by selecting the most representative atoms from the extended dictionary that are correlated to the similar signal set. In the LAJSR framework, the selected similar signals are simultaneously represented by the local adaptive dictionary, and the obtained sparse representation coefficients are further weighted by a sparsity concentration index vector which aims to concentrate and highlight the coefficients on the expected class. Experimental results on two benchmark hyperspectral data sets have demonstrated that the proposed LAJSR method is much more effective than existing JSR and SVM methods, especially in the case of small sample sizes.

Visual tracking via context-aware local sparse appearance model

Most existing local sparse trackers are prone to drifting away as they do not make use of discriminative information of local patches. In this paper, we propose an effective context-aware local sparse appearance model to alleviate the drift problem caused by background clutter and occlusions. First, considering that different local patches should have different impacts on the likelihood computation, we present a novel Impact Allocation Strategy (IAS) with integration of the spatial-temporal context. Varying positive impact factors are adaptively assigned to different local patches based on their ability distinguishing the spatial context, which provides discriminative information to prevent the tracker from drifting. Furthermore, we exploit temporal context to introduce some historical information for more accurate locating. Second, we present a new patch-based dictionary update method being able to update each patch independently with the validation of effectiveness. On the one hand, we introduce sparsity concentration index to check whether the local patch to be updated is a valid local patch from the target object. On the other hand, spatial context is further employed to eliminate the effect of the background. Experimental results show the superiority and competitiveness of the proposed method on the benchmark data set compared to other state-of-the-art algorithms.

Joint compressive representation for multi-feature tracking

Abstract Multi-feature tracking is an interesting but challenging task. Many previous works just consider the fusion of different features from identical spectral image or identical features from different spectral images alone, which makes them be quite distinct from each other and be difficult to be integrated naturally. To address this problem, we propose a unified multi-feature tracking framework based on joint compressive sensing in this paper. Our framework can accept features extracted from identical spectral or different spectral images, and provide the flexibility to arbitrary add or remove feature. We formulate the multi-feature tracking as a joint minimization problem of multiple l1-norms with inequality constraint of multiple l2-norms, and derive a customized augmented Lagrange multiplier algorithm to solve the minimization problem, which provides efficient object tracking with both low computational burden and high accuracy. Besides, a collaborative template update scheme induced by sparsity concentration index is developed to keep track of the most representative templates throughout the tracking procedure. Experimental results on challenging visible and infrared sequences clearly demonstrate the robustness and effectiveness of the proposed method.

Classification of machinery vibration signals based on group sparse representation

The working condition of mechanical equipment can be reflected by vibration signals collected from it. Accurate classification of these vibration signals is helpful for the machinery fault diagnosis. In recent years, the L1-norm regularization based sparse representation for classification (SRC) has obtained huge success in image recognition, especially in face recognition. However, the investigation of SRC for machinery vibration signals shows that the accuracy and sparsity concentration index are not high enough. In this paper, a new classification method for machinery vibration signals is proposed, in which the L1L2-norm regularization based sparse representation, i.e. group sparse representation, is recommended as a coding strategy. The method achieves its idea classification performance by three steps. Firstly, time-domain vibration signals, including training and test samples, are transformed to frequency-domain to reduce the influence of corrupting noise. Then, the transform coefficient vectors of the test samples are coded with a combination of L1-norm and L2-norm constrain on a dictionary, which is constructed by merging the transform coefficient vectors of the training samples. At last, the fault types of the test samples are labeled by identifying their minimal reconstruction errors. The classification results of simulated and experimental vibration signals demonstrate the superiority of proposed method in comparison with the state-of-the-art classifiers.

Iris recognition based on robust principal component analysis

Iris images acquired under different conditions often suffer from blur, occlusion due to eyelids and eyelashes, specular reflection, and other artifacts. Existing iris recognition systems do not perform well on these types of images. To overcome these problems, we propose an iris recognition method based on robust principal component analysis. The proposed method decomposes all training images into a low-rank matrix and a sparse error matrix, where the low-rank matrix is used for feature extraction. The sparsity concentration index approach is then applied to validate the recognition result. Experimental results using CASIA V4 and IIT Delhi V1iris image databases showed that the proposed method achieved competitive performances in both recognition accuracy and computational efficiency.

Hyperspectral Image Classification Using Kernel Sparse Representation and Semilocal Spatial Graph Regularization

This letter presents a postprocessing algorithm for a kernel sparse representation (KSR)-based hyperspectral image classifier, which is based on the integration of spatial and spectral information. A pixelwise KSR is first used to find the sparse coefficient vectors of the hyperspectral image. Then, a sparsity concentration index (SCI) rule-guided semilocal spatial graph regularization (SSG), called SSG+SCI, is proposed to determine refined sparse coefficient vectors that promote spatial continuity within each class. Finally, these refined coefficient vectors are used to obtain the final classification map. Compared with previous approaches based on similar spatial-spectral postprocessing strategies, SSG+SCI clearly outperforms their results in terms of accuracy and the number of training samples, as it is demonstrated with two real hyperspectral images.

Sparse representation for gastropathy disease diagnosis

Automatic diagnosis of gastropathy disease plays an important role in supporting the clinician and speeding up the examination process. In this paper, we propose a general algorithm for stomach disease detection using sparse representation, which can get robust result even using a small size of high-dimensional training dataset. First, a noval large scale dictionary selection model based on the low rank property is designed to initialize the meaningful dictionary with minimal size. Then in comparison with the state-of-the-arts, we consider disease diagnosing as both supervised classification issue and one-class classification issue, where a new criterion, weighted sparsity concentration index (WSCI) is designed for outlier detection. Experiments on our dataset and comparisons with state-of-the-arts demonstrate the effectiveness and effciency of our algorithm.