Multi-task Sparse Representation Research Articles

Learning Multi-Task Sparse Representation Based on Fisher Information

Multi-task learning deals with multiple related tasks simultaneously by sharing knowledge. In a typical deep multi-task learning model, all tasks use the same feature space and share the latent knowledge. If the tasks are weakly correlated or some features are negatively correlated, sharing all knowledge often leads to negative knowledge transfer among. To overcome this issue, this paper proposes a Fisher sparse multi-task learning method. It can obtain a sparse sharing representation for each task. In such a way, tasks share features on a sparse subspace. Our method can ensure that the knowledge transferred among tasks is beneficial. Specifically, we first propose a sparse deep multi-task learning model, and then introduce Fisher sparse module into traditional deep multi-task learning to learn the sparse variables of task. By alternately updating the neural network parameters and sparse variables, a sparse sharing representation can be learned for each task. In addition, in order to reduce the computational overhead, an heuristic method is used to estimate the Fisher information of neural network parameters. Experimental results show that, comparing with other methods, our proposed method can improve the performance for all tasks, and has high sparsity in multi-task learning.

Multitask Sparse Representation of Two-Dimensional Variational Mode Decomposition Components for SAR Target Recognition

A synthetic aperture radar (SAR) automatic target recognition (ATR) method is developed based on the two-dimensional variational mode decomposition (2D-VMD). 2D-VMD decomposes original SAR images into multiscale components, which depict the time-frequency properties of the targets. The original image and its 2D-VMD components are highly correlated, so the multitask sparse representation is chosen to jointly represent them. According to the resulted reconstruction errors of different classes, the target label of test sample can be classified. The moving and stationary target acquisition and recognition (MSTAR) dataset is used to set up the standard operating condition (SOC) and several extended operating conditions (EOCs) including configuration variants, depression angle variances, noise corruption, and partial occlusion to test and validate the proposed method. The results confirm the effectiveness and robustness of the proposed method compared with several state-of-the-art SAR ATR references.

Recognition of Power Equipment Based on Multitask Sparse Representation

Image analysis of power equipment has important practical significance for power-line inspection and maintenance. This paper proposes an image recognition method for power equipment based on multitask sparse representation. In the feature extraction stage, based on the two-dimensional (2D) random projection algorithm, multiple projection matrices are constructed to obtain the multilevel features of the image. In the classification process, considering that the image acquisition process will inevitably be affected by factors such as light conditions and noise interference, the proposed method uses the multitask compressive sensing algorithm (MtCS) to jointly represent multiple feature vectors to improve the accuracy and robustness of reconstruction. In the experiment, the images of three types of typical power equipment of insulators, transformers, and circuit breakers are classified. The correct recognition rate of the proposed method reaches 94.32%. In addition, the proposed method can maintain strong robustness under the conditions of noise interference and partial occlusion, which further verifies its effectiveness.

A Multiview SAR Target Recognition Method Using Inner Correlation Analysis.

This paper proposes a synthetic aperture radar (SAR) image target recognition method using multiple views and inner correlation analysis. Due to the azimuth sensitivity of SAR images, the inner correlation between multiview images participating in recognition is not stable enough. To this end, the proposed method first clusters multiview SAR images based on image correlation and nonlinear correlation information entropy (NCIE) in order to obtain multiple view sets with strong internal correlations. For each view set, the multitask sparse representation is used to reconstruct the SAR images in it to obtain high-precision reconstructions. Finally, the linear weighting method is used to fuse the reconstruction errors from different view sets and the target category is determined according to the fusion error. In the experiment, the tests are conducted based on the MSTAR dataset, and the results validate the effectiveness of the proposed method.

Reduce Surface Electromyography Channels for Gesture Recognition by Multitask Sparse Representation and Minimum Redundancy Maximum Relevance.

Surface electromyography- (sEMG-) based gesture recognition is widely used in rehabilitation training, artificial prosthesis, and human-computer interaction. The purpose of this study is to simplify the sEMG devices by reducing channels while achieving comparably high gesture recognition accuracy. We propose a compound channel selection scheme by combining the variable selection algorithms based on multitask sparse representation (MTSR) and minimum Redundancy Maximum Relevance (mRMR). Specifically, channelwise features are first extracted to compose channel-feature paired variables, for which variable selection procedures by MTSR and mRMR are carried out, respectively. Then, we rank all the channels according to their occurrences in each variable selection procedure and figure out a certain number of informative channels by fusing these rankings of channels. Finally, the gesture classification performance using the selected channels is evaluated by the support vector machine (SVM) classifier. Experiment results validate the effectiveness of this proposed method.

Target Recognition of SAR Images Based on Complex Bidimensional Empirical Mode Decomposition

A target recognition method for synthetic aperture radar (SAR) image based on complex bidimensional empirical mode decomposition (C-BEMD) is proposed. C-BEMD is used to decompose the original SAR image to obtain multilevel complex bidimensional intrinsic mode functions (BIMF), which reflect the two-dimensional time-frequency characteristics of the target. In the classification stage, the decomposed multilevel BIMFs are represented using the multitask sparse representation. Finally, the target category of the test sample is determined according to the reconstruction errors related to different training classes. In the experiment, the standard operating condition (SOC) and extended operating conditions (EOC) are designed based on the MSTAR dataset to test and verify the proposed method. The results confirm the effectiveness and robustness of the method.

A temporal sparse collaborative appearance model for visual tracking

This paper proposes a temporal sparse collaborative appearance model in the particle filter frame for constructing a robust visual tracker, denoted by TSCAM tracker. The proposed collaborative appearance model includes two important branches: temporal Laplacian multi-task discriminative branch and similarity branch. The model in the first branch develops the classical Laplacian multi-task reverse sparse representation model by adding a temporality regularization term in the appearance model, which can get more global feature information and capture the relation of the adjacent frames. The application of sparse generative model (SGM) in the second branch can exploit the local appearance information of patches and occlusion to moderate the appearance variation and eliminate the partial occlusion. Therefore, with the alliance of above two branches, our proposed TSCAM tracker can track the object with occlusion and appearance variations effectively. Numerous experiments on various challenging videos of some databases illustrate that our proposed TSCAM tracker performs favorably against several state-of-the-art trackers.

Hyperspectral Image Target Detection via Weighted Joint K-Nearest Neighbor and Multitask Learning Sparse Representation

Multitask sparse representation method improves the detection performance by constructing multiple associated sub-sparse representation tasks and jointly learning multiple sub-sparse representation tasks, and this method can make use of the spectral information. However, the using of spatial information needs to be improved. This paper designs a hyperspectral image target detection method which can both make use of spectral and spatial information, that is a weighted joint k-nearest neighbor and multitask learning sparse representation method (WJNN-MTL-SR) is proposed. This method mainly consists of the following steps:1) using multitask sparse representation to obtain the representation residuals. 2) weighted joint k-nearest neighbor is used into the joint region of test pixels to obtain the weighted joint Euclidean distance. 3) a decision function, combining the weighted joint Euclidean distance and residuals of the multitask sparse representation, is used to get target detection result. Experimental results demonstrate that the proposed method show better detection performance than state-of-the-art methods.

Multi-task hybrid dictionary learning for vehicle classification in sensor networks

In this article, we propose a novel multi-task hybrid dictionary learning approach for moving vehicle classification tasks using multi-sensor networks to improve the classification accuracy in complex scenes with low time complexity, which considers both correlations and complementary information among multiple heterogeneous sensors simultaneously to learn a hybrid dictionary within observations of each sensor. The efficient hybrid dictionary consists of a synthesis dictionary and an analysis dictionary, where discriminative codes can be generated by the trained analysis dictionary and class-specific discriminative reconstruction can be achieved by the trained synthesis dictionary. Extensive experiments are conducted on real data sets captured by the multiple heterogeneous sensors, and the results demonstrate that the proposed method can use the multi-feature fusion method to improve the vehicle classification accuracy, and it can learn a hybrid dictionary to make sure that the sparse coding matrix is obtained by simple linear mapping function. Moreover, the problem of [Formula: see text]-norm[Formula: see text] sparse coding can been solved, to reduce the time complexity of this algorithm, compared with support vector machine, sparse representation classification, label consistent KSVD, Fisher discrimination dictionary learning, hybrid dictionary learning, multi-task sparse representation classification, and multi-task Fisher discrimination dictionary learning algorithms.

Radar Target Recognition Using Salient Keypoint Descriptors and Multitask Sparse Representation

In this paper, we propose a novel approach to recognize radar targets on inverse synthetic aperture radar (ISAR) and synthetic aperture radar (SAR) images. This approach is based on the multiple salient keypoint descriptors (MSKD) and multitask sparse representation based classification (MSRC). Thus, to characterize the targets in the radar images, we combine the scale-invariant feature transform (SIFT) and the saliency map. The purpose of this combination is to reduce the number of SIFT keypoints by keeping only those located in the target area (salient region); this speeds up the recognition process. After that, we compute the feature vectors of the resulting salient SIFT keypoints (MSKD). This methodology is applied for both training and test images. The MSKD of the training images leads to constructing the dictionary of a sparse convex optimization problem. To achieve the recognition, we adopt the MSRC taking into consideration each vector in the MSKD as a task. This classifier solves the sparse representation problem for each task over the dictionary and determines the class of the radar image according to all sparse reconstruction errors (residuals). The effectiveness of the proposed approach method has been demonstrated by a set of extensive empirical results on ISAR and SAR images databases. The results show the ability of the proposed method to predict adequately the aircraft and the ground targets.

Multiple 3-D Feature Fusion Framework for Hyperspectral Image Classification

Due to the 3-D nature of hyperspectral images, as well as the spatial properties (such as regularity and continuity) of land covers, many 3-D feature extraction operators have been designed to fully exploit the joint spatial–spectral information. However, the large amount of obtained features can suffer from the “curse of dimensionality” problem, especially for the small training sample set. Moreover, various spatial–spectral features can represent the characteristics of the hyperspectral image from different aspects. In this paper, a multiple 3-D feature fusion framework (M3DF3) has been proposed for hyperspectral image classification. First, we extend the 2-D Gabor surface feature into 3-D (3DSF) domains to comply with the spatial–spectral structure of the hyperspectral image, which is directly applied on the original hyperspectral image instead of the Gabor features. Second, three 3-D feature extraction methods, including the 3-D morphological profile, the 3-D local binary pattern, and the proposed 3DSF, that, respectively, characterize the hyperspectral image from three different angles, i.e., morphology, local dependence, and shape smoothness, are fused under a multitask sparse representation framework to take full advantage of the multiple 3-D features together. The proposed M3DF3 approach was fully tested on three real-world hyperspectral image data, i.e., the widely used Indian Pines, Pavia University, and Houston University. The results show that our method can achieve as high as 68.22%, 79.44%, and 72.84% accuracies, respectively, even when only few samples, i.e., three samples per class, are used for training.

Face recognition from unconstrained three-dimensional face images using multitask sparse representation

We propose and evaluate a three-dimensional (3D) face recognition approach that applies the speeded up robust feature (SURF) algorithm to the depth representation of shape index map, under real-world conditions, using only a single gallery sample for each subject. First, the 3D scans are preprocessed, then SURF is applied on the shape index map to find interest points and their descriptors. Each 3D face scan is represented by keypoints descriptors, and a large dictionary is built from all the gallery descriptors. At the recognition step, descriptors of a probe face scan are sparsely represented by the dictionary. A multitask sparse representation classification is used to determine the identity of each probe face. The feasibility of the approach that uses the SURF algorithm on the shape index map for face identification/authentication is checked through an experimental investigation conducted on Bosphorus, University of Milano Bicocca, and CASIA 3D datasets. It achieves an overall rank one recognition rate of 97.75%, 80.85%, and 95.12%, respectively, on these datasets.

Local feature approach to dorsal hand vein recognition by Centroid-based Circular Key-point Grid and fine-grained matching

Due to the great progress made by local feature matching in both performance and robustness of dorsal hand vein recognition, this paper proposes a novel and effective approach for such an issue by improving two major steps of the SIFT-like framework, i.e. key-point detection and matching. For the former, a new key-point generation pattern, namely Centroid-based Circular Key-point Grid (CCKG), is presented, which efficiently localizes a certain number of points on the dorsal hand for the following SIFT feature extraction, leading to a discriminative description. In contrast to the existing key-point detectors, CCKG comprehensively accounts for the properties of the dorsal hand, including the vein network as well as the surrounding corium region, and hence achieves both good representativeness and low complexity. For the latter, a fine-grained matching process is introduced which makes use of Multi-task Sparse Representation Classifier (MtSRC). Compared with the traditional coarse-grained one that counts the number of associated SIFT features between the gallery and probe dorsal hand images, MtSRC precisely calculates the error of each feature of the probe as reconstructed by the gallery features, and all the errors of the probe features are combined for similarity measurement, reaching a better accuracy in recognition. The proposed approach is evaluated on the NCUT Part A database and shows its effectiveness in both the identification and verification scenarios. Additionally, the experimental results achieved on the NCUT Part B dataset highlight its generality and robustness to low quality images.

Face Recognition by Exploiting Local Gabor Features With Multitask Adaptive Sparse Representation

For a human face, the Gabor transform can extract its multiple scale and orientation features that are very useful for the recognition. In this paper, the Gabor-feature-based face recognition is formulated as a multitask sparse representation model, in which the sparse coding of each Gabor feature is regarded as a task. To effectively exploit the complementary yet correlated information among different tasks, a flexible representation algorithm termed multitask adaptive sparse representation (MASR) is proposed. The MASR algorithm not only restricts Gabor features of one test sample to be jointly represented by training atoms from the same class but also promotes the selected atoms for these features to be varied within each class, thus allowing better representation. In addition, to use the local information, we operate the MASR on local regions of Gabor features. Then, by considering the structural characteristics of the face and the effects of the external interferences, a structural-residual weighting strategy is proposed to adaptively fuse the decision of each region. Experiments on various datasets verify the effectiveness of the proposed method in dealing with face occlusion, corruption, small number of training samples, as well as variations of lighting and expression.

Towards 3D Face Recognition in the Real: A Registration-Free Approach Using Fine-Grained Matching of 3D Keypoint Descriptors

Registration algorithms performed on point clouds or range images of face scans have been successfully used for automatic 3D face recognition under expression variations, but have rarely been investigated to solve pose changes and occlusions mainly since that the basic landmarks to initialize coarse alignment are not always available. Recently, local feature-based SIFT-like matching proves competent to handle all such variations without registration. In this paper, towards 3D face recognition for real-life biometric applications, we significantly extend the SIFT-like matching framework to mesh data and propose a novel approach using fine-grained matching of 3D keypoint descriptors. First, two principal curvature-based 3D keypoint detectors are provided, which can repeatedly identify complementary locations on a face scan where local curvatures are high. Then, a robust 3D local coordinate system is built at each keypoint, which allows extraction of pose-invariant features. Three keypoint descriptors, corresponding to three surface differential quantities, are designed, and their feature-level fusion is employed to comprehensively describe local shapes of detected keypoints. Finally, we propose a multi-task sparse representation based fine-grained matching algorithm, which accounts for the average reconstruction error of probe face descriptors sparsely represented by a large dictionary of gallery descriptors in identification. Our approach is evaluated on the Bosphorus database and achieves rank-one recognition rates of 96.56, 98.82, 91.14, and 99.21 % on the entire database, and the expression, pose, and occlusion subsets, respectively. To the best of our knowledge, these are the best results reported so far on this database. Additionally, good generalization ability is also exhibited by the experiments on the FRGC v2.0 database.

Object Tracking via Robust Multitask Sparse Representation

Sparse representation has been applied to the object tracking problem. Mining the self-similarities between particles via multitask learning can improve tracking performance. However, some particles may be different from others when they are sampled from a large region. Imposing all particles share the same structure may degrade the results. To overcome this problem, we propose a tracking algorithm based on robust multitask sparse representation (RMTT) in this letter. When we learn the particle representations, we decompose the sparse coefficient matrix into two parts in our algorithm. Joint sparse regularization is imposed on one coefficient matrix while element-wise sparse regularization is imposed on another matrix. The former regularization exploits self-similarities of particles while the later one considers the differences between them. Experiments on the benchmark data show the superior performance over other state-of-art algorithms.