Subspace Learning Method Research Articles

Extracting shared subspace incrementally for multi-label image classification

With the popularity of internet technology, thousands of new images with multiple labels appear on the web every day. For a large number of images updated daily on the websites, it is of ever-increasing importance to classify these new multi-label images online in real time. Accordingly, this paper presents an incremental shared subspace learning method for multi-label image classification. With the incremental lossless matrix factorization, the proposed algorithm can be incrementally performed without using original existing input data, thus high computational complexity involved in extracting the shared subspace can be avoided. Several publicly available multi-label image datasets are used to evaluate the proposed method. Experimental results demonstrate that the proposed approach is much more efficient than the non-incremental methods without decreasing the classification performance.

CLASSIFICATION OF POLARIMETRIC SAR IMAGE BASED ON THE SUBSPACE METHOD

Abstract. Land cover classification is one of the most significant applications in remote sensing. Compared to optical sensing technologies, synthetic aperture radar (SAR) can penetrate through clouds and have all-weather capabilities. Therefore, land cover classification for SAR image is important in remote sensing. The subspace method is a novel method for the SAR data, which reduces data dimensionality by incorporating feature extraction into the classification process. This paper uses the averaged learning subspace method (ALSM) method that can be applied to the fully polarimetric SAR image for classification. The ALSM algorithm integrates three-component decomposition, eigenvalue/eigenvector decomposition and textural features derived from the gray-level cooccurrence matrix (GLCM). The study site, locates in the Dingxing county, in Hebei Province, China. We compare the subspace method with the traditional supervised Wishart classification. By conducting experiments on the fully polarimetric Radarsat-2 image, we conclude the proposed method yield higher classification accuracy. Therefore, the ALSM classification method is a feasible and alternative method for SAR image.

Sparse tensor discriminant analysis

The classical linear discriminant analysis has undergone great development and has recently been extended to different cases. In this paper, a novel discriminant subspace learning method called sparse tensor discriminant analysis (STDA) is proposed, which further extends the recently presented multilinear discriminant analysis to a sparse case. Through introducing the L1 and L2 norms into the objective function of STDA, we can obtain multiple interrelated sparse discriminant subspaces for feature extraction. As there are no closed-form solutions, k-mode optimization technique and the L1 norm sparse regression are combined to iteratively learn the optimal sparse discriminant subspace along different modes of the tensors. Moreover, each non-zero element in each subspace is selected from the most important variables/factors, and thus STDA has the potential to perform better than other discriminant subspace methods. Extensive experiments on face databases (Yale, FERET, and CMU PIE face databases) and the Weizmann action database show that the proposed STDA algorithm demonstrates the most competitive performance against the compared tensor-based methods, particularly in small sample sizes.

RETRACTED ARTICLE: Sparse tensor CCA for color face recognition

This paper proposes a subspace learning method, named as sparse tensor canonical correlation analysis (ST-CCA), for color face recognition. A sample image is formalized as high-order tensors to preserve the inherent structure of the color face images. We utilize sparse canonical correlation analysis (SCCA) to choose gene. For each pair of tensors, SCCA generates the sparse loadings alternately, which is helpful for choosing significant variables to reduce dimensions and eliminate the redundancies of tensors. We use the elastic net as constraint condition to attack the collinearity problem by decorrelating and selecting the sufficient variables irrespective of the limited dimensions. Furthermore, ST-CCA gains stable recognition rates because the alternating least square solution converges. ST-CCA is convex with different initials of the projection matrices. Experimental results on AR face database and LFW face database show the superior performance of our method over the state-of-the-art ones.

Learning a subspace for clustering via pattern shrinking

Clustering is a basic technique in information processing. Traditional clustering methods, however, are not suitable for high dimensional data. Thus, learning a subspace for clustering has emerged as an important research direction. Nevertheless, the meaningful data are often lying on a low dimensional manifold while existing subspace learning approaches cannot fully capture the nonlinear structures of hidden manifold. In this paper, we propose a novel subspace learning method that not only characterizes the linear and nonlinear structures of data, but also reflects the requirements of following clustering. Compared with other related approaches, the proposed method can derive a subspace that is more suitable for high dimensional data clustering. Promising experimental results on different kinds of data sets demonstrate the effectiveness of the proposed approach.

Singing speaker clustering based on subspace learning in the GMM mean supervector space

In this study, we propose algorithms based on subspace learning in the GMM mean supervector space to improve performance of speaker clustering with speech from both reading and singing. As a speaking style, singing introduces changes in the time-frequency structure of a speaker’s voice. The purpose of this study is to introduce advancements for speech systems such as speech indexing and retrieval which improve robustness to intrinsic variations in speech production. Speaker clustering techniques such as k-means and hierarchical are explored for analysis of acoustic space differences of a corpus consisting of reading and singing of lyrics for each speaker. Furthermore, a distance based on fuzzy c-means membership degrees is proposed to more accurately measure clustering difficulty or speaker confusability. Two categories of subspace learning methods are studied: unsupervised based on LPP, and supervised based on PLDA. Our proposed clustering method based on PLDA is a two stage algorithm: where first, initial clusters are obtained using full dimension supervectors, and next, each cluster is refined in a PLDA subspace resulting in a more speaker dependent representation that is less sensitive to speaking style. It is shown that LPP improves average clustering accuracy by 5.1% absolute versus a hierarchical baseline for a mixture of reading and singing, and PLDA based clustering increases accuracy by 9.6% absolute versus a k-means baseline. The advancements offer novel techniques to improve model formulation for speech applications including speaker ID, audio search, and audio content analysis.

Efficient optimally regularized discriminant analysis

Regularized discriminant analysis (RDA) and its special case uncorrelated linear discriminant analysis (ULDA) are important subspace learning methods proposed recently to handle the small sample size (SSS) problem of linear discriminant analysis (LDA). One important unsolved issue of RDA is how to automatically determine an appropriate regularization parameter without resorting to unscalable procedures like cross-validation (CV). In this paper, we develop a novel efficient algorithm to automatically estimate the regularization parameter based on a geometric interpretation of RDA. We further provide a formal analysis of the proposed method, and show that it is robust to the perturbation in the feature space of the training data. The extensive experiments on various benchmark datasets verify the scalability and effectiveness of our approach, compared with the state-of-the-art algorithms.

Biview face recognition in the shape–texture domain

Face recognition is one of the biometric identification methods with the highest potential. The existing face recognition algorithms relying on the texture information of face images are affected greatly by the variation of expression, scale and illumination. Whereas the algorithms based on the shape topology weaken the influence of illumination to some extent, but the impact of expression, scale and illumination on face recognition is still unsolved. To this end, we propose a new method for face recognition by integrating texture information with shape information, called biview face recognition algorithm. The texture models are constructed by using subspace learning methods and shape topologies are formed by building graphs for face images. The proposed biview face recognition method is compared with recognition algorithms merely based on texture or shape information. Experimental results of recognizing faces under the variation of illumination, expression and scale demonstrate that the performance of the proposed biview face recognition outperforms texture-based and shape-based algorithms.

Generalized N-dimensional independent component analysis and its application to multiple feature selection and fusion for image classification

We propose a multilinear independent component analysis (ICA) framework called generalized N-dimensional ICA (GND-ICA) by extending the conventional linear ICA based on the multilinear algebra. Unlike the linear ICA that only treats one-dimensional data, the proposed GND-ICA treats N-dimensional data as a tensor without any preprocess of data vectorization. We furthermore introduce two types of GND-ICA solutions and analyze their efficiency and effectiveness. As an application, the GND-ICA can be used for multiple feature fusion and representation for color image classification. Many features extracted from a given image are constructed as a tensor. The feature tensor can be effective represented by GND-ICA. Compared with the conventional linear subspace learning methods, GND-ICA is capable of obtaining more distinctive representation for color image classification.

Supervised and Unsupervised Parallel Subspace Learning for Large-Scale Image Recognition

Subspace learning is an effective and widely used image feature extraction and classification technique. However, for the large-scale image recognition issue in real-world applications, many subspace learning methods often suffer from large computational burden. In order to reduce the computational time and improve the recognition performance of subspace learning technique under this situation, we introduce the idea of parallel computing which can reduce the time complexity by splitting the original task into several subtasks. We develop a parallel subspace learning framework. In this framework, we first divide the sample set into several subsets by designing two random data division strategies that are equal data division and unequal data division. These two strategies correspond to equal and unequal computational abilities of nodes under parallel computing environment. Next, we calculate projection vectors from each subset in parallel. The graph embedding technique is employed to provide a general formulation for parallel feature extraction. After combining the extracted features from all nodes, we present a unified criterion to select most distinctive features for classification. Under the developed framework, we separately propose supervised and unsupervised parallel subspace learning approaches, which are called parallel linear discriminant analysis (PLDA) and parallel locality preserving projection (PLPP). PLDA selects features with the largest Fisher scores by estimating the weighted and unweighted sample scatter, while PLPP selects features with the smallest Laplacian scores by constructing a whole affinity matrix. Theoretically, we analyze the time complexities of proposed approaches and provide the fundamental supports for applying random division strategies. In the experiments, we establish two real parallel computing environments and employ four public image and video databases as the test data. Experimental results demonstrate that the proposed approaches outperform several related supervised and unsupervised subspace learning methods, and significantly reduce the computational time.

Locality-Based Discriminant Neighborhood Embedding

In this article, we develop a linear supervised subspace learning method called locality-based discriminant neighborhood embedding (LDNE), which can take advantage of the underlying submanifold-based structures of the data for classification. Our LDNE method can simultaneously consider both ‘locality’ of locality preserving projection (LPP) and ‘discrimination’ of discriminant neighborhood embedding (DNE) in manifold learning. It can find an embedding that not only preserveslocalinformationtoexploretheintrinsicsubmanifoldstructureofdatafromthesameclass, but also enhances the discrimination among submanifolds from different classes. To investigate the performance of LDNE, we compare it with the state-of-the-art dimensionality reduction techniques such as LPP and DNE on publicly available datasets. Experimental results show that our LDNE can be an effective and robust method for classification.

Palmprint Recognition across Different Devices

In this paper, the problem of Palmprint Recognition Across Different Devices (PRADD) is investigated, which has not been well studied so far. Since there is no publicly available PRADD image database, we created a non-contact PRADD image database containing 12,000 grayscale captured from 100 subjects using three devices, i.e., one digital camera and two smart-phones. Due to the non-contact image acquisition used, rotation and scale changes between different images captured from a same palm are inevitable. We propose a robust method to calculate the palm width, which can be effectively used for scale normalization of palmprints. On this PRADD image database, we evaluate the recognition performance of three different methods, i.e., subspace learning method, correlation method, and orientation coding based method, respectively. Experiments results show that orientation coding based methods achieved promising recognition performance for PRADD.

Twin support vector machines and subspace learning methods for microcalcification clusters detection

This paper presents a novel framework for microcalcification clusters (MCs) detection in mammograms. The proposed framework has three main parts: (1) first, MCs are enhanced by using a simple-but-effective artifact removal filter and a well-designed high-pass filter; (2) thereafter, subspace learning algorithms can be embedded into this framework for subspace (feature) selection of each image block to be handled; and (3) finally, in the resulted subspaces, the MCs detection procedure is formulated as a supervised learning and classification problem, and in this work, the twin support vector machine (TWSVM) is developed in decision-making of MCs detection. A large number of experiments are carried out to evaluate and compare the MCs detection approaches, and the effectiveness of the proposed framework is well demonstrated.

Discriminant sparse neighborhood preserving embedding for face recognition

Sparse subspace learning has drawn more and more attentions recently. However, most of the sparse subspace learning methods are unsupervised and unsuitable for classification tasks. In this paper, a new sparse subspace learning algorithm called discriminant sparse neighborhood preserving embedding (DSNPE) is proposed by adding the discriminant information into sparse neighborhood preserving embedding (SNPE). DSNPE not only preserves the sparse reconstructive relationship of SNPE, but also sufficiently utilizes the global discriminant structures from the following two aspects: (1) maximum margin criterion (MMC) is added into the objective function of DSNPE; (2) only the training samples with the same label as the current sample are used to compute the sparse reconstructive relationship. Extensive experiments on three face image datasets (Yale, Extended Yale B and AR) demonstrate the effectiveness of the proposed DSNPE method.

Image classification by multimodal subspace learning

In recent years we witnessed a surge of interest in subspace learning for image classification. However, the previous methods lack of high accuracy since they do not consider multiple features of the images. For instance, we can represent a color image by finding a set of visual features to represent the information of its color, texture and shape. According to the “Patch Alignment” Framework, we developed a new subspace learning method, termed Semi-Supervised Multimodal Subspace Learning (SS-MMSL), in which we can encode different features from different modalities to build a meaningful subspace. In particular, the new method adopts the discriminative information from the labeled data to construct local patches and aligns these patches to get the optimal low dimensional subspace for each modality. For local patch construction, the data distribution revealed by unlabeled data is utilized to enhance the subspace learning. In order to find a low dimensional subspace wherein the distribution of each modality is sufficiently smooth, SS-MMSL adopts an alternating and iterative optimization algorithm to explore the complementary characteristics of different modalities. The iterative procedure reaches the global minimum of the criterion due to the strong convexity of the criterion. Our experiments of image classification and cartoon retrieval demonstrate the validity of the proposed method.

Fisher Difference Discriminant Analysis: Determining the Effective Discriminant Subspace Dimensions for Face Recognition

In most manifold learning based subspace discriminant analysis algorithms, how to construct the local neighborhood graphs and determine the effective discriminant subspace dimensions in applications are difficult but important problems. In this paper, we propose a novel supervised subspace learning method called Fisher Difference Discriminant Analysis (FDDA) for linear dimensionality reduction. FDDA introduces the local soft scatter to characterize the distributions of the data set. By combining Fisher criterion and difference criterion together, FDDA obtains the optimal discriminant subspace, on which a large margin between different classes is provided for classification. Eigenvalue analysis shows that the effective discriminant subspace dimensions of FDDA can be automatically determined by the number of positive eigenvalues and are robust to noise and invariant to rotations, rescalings and translations of the data. Comprehensive comparison and extensive experiments show that FDDA is superior to some state-of-the-art techniques in face recognition.

Background modeling by subspace learning on spatio-temporal patches

This paper presents a novel background model for video surveillance—Spatio-Temporal Patch based Background Modeling (STPBM). We use spatio-temporal patches, called bricks, to characterize both the appearance and motion information. Our method is based on the observation that all the background bricks at a given location under all possible lighting conditions lie in a low dimensional background subspace, while bricks with moving foreground are widely distributed outside. An efficient online subspace learning method is presented to capture the subspace, which is able to model the illumination changes more robustly than traditional pixel-wise or block-wise methods. Experimental results demonstrate that the proposed method is insensitive to drastic illumination changes yet capable of detecting dim foreground objects under low contrast. Moreover, it outperforms the state-of-the-art in various challenging scenes with illumination changes.

Dimensionality Reduction Based on Neighborhood Preserving and Marginal Discriminant Embedding

Data with high dimensionality often occurs, which will produce large time and energy overheads when directly used in classification tasks. So, as one of the most important fields in machine learning, dimensionality reduction has been paid more and more attention and has achieved a prodigious progress in the theory and algorithm research. Linear Graph embedding (LGE) model is an efficient tool for dimensionality reduction. According to the problems of supervised dimensionality reduction with Non-Gaussian data distributions and at the same time consider neighborhood preserving relations among samples, a novel subspace learning method, neighborhood preserving and marginal discriminant embedding (NP-MDE), is proposed based on LGE and marginal Fisher analysis in this paper. NP-MDE could minimize the within-class scatter and meanwhile maximize the margin among different classes. Moreover, the neighborhood structure with each class is preserved. Experiments on Yale face image data sets show that after dimensionality reduction using NP-MDE, the average classification accuracy is very good.

Block principal component analysis with L1-norm for image analysis

Block principal component analysis (BPCA) is an important subspace learning method in modern image analysis. The utilization of the L2-norm, however, makes it sensitive to outliers. In this paper, we propose an L1-norm-based BPCA (BPCA-L1) as a robust alternative to BPCA. We show the equivalence between the L1-norm-based two-dimensional principal component analysis (2DPCA-L1) and the L1-norm-based principal component analysis (PCA-L1), both of which can be formulated as special cases of BPCA-L1. Experiments of image reconstruction and classification on benchmark image sets show the effectiveness of the proposed method.

Mining Visual Collocation Patterns via Self-Supervised Subspace Learning

Traditional text data mining techniques are not directly applicable to image data which contain spatial information and are characterized by high-dimensional visual features. It is not a trivial task to discover meaningful visual patterns from images because the content variations and spatial dependence in visual data greatly challenge most existing data mining methods. This paper presents a novel approach to coping with these difficulties for mining visual collocation patterns. Specifically, the novelty of this work lies in the following new contributions: 1) a principled solution to the discovery of visual collocation patterns based on frequent itemset mining and 2) a self-supervised subspace learning method to refine the visual codebook by feeding back discovered patterns via subspace learning. The experimental results show that our method can discover semantically meaningful patterns efficiently and effectively.