Yibin Tang

Design of Sparse FIR Filters With Joint Optimization of Sparsity and Filter Order

In this paper, two novel algorithms are developed to design sparse linear-phase (LP) FIR filters. Compared to traditional design methods, they can jointly optimize coefficient sparsity and order of an LP FIR filter, so as to achieve a balance between filtering performance and implementation efficiency. The design problem under consideration is formally cast as a regularized l0-norm minimization problem, which is then tackled by two different design algorithms. In the first proposed algorithm, the objective function of the original design problem is replaced by its upper bound, which leads to a weighted l0-norm minimization problem, while in the second one a group of auxiliary variables are introduced such that the original design problem can be equivalently transformed to another weighted l0-norm minimization problem. The iterative-reweighted-least-squares (IRLS) algorithm is employed with appropriate modifications to solve both weighted l0-norm minimization problems. Simulation results show that, compared to traditional approaches, the proposed algorithms can achieve comparable or better design results in terms of both sparsity and effective filter order.

Image denoising via sparse coding using eigenvectors of graph Laplacian

Image denoising plays an important role in image processing, which aims to separate clean images from the noisy images. A number of methods have been presented to deal with this practical problem in the past decades. In this paper, a sparse coding algorithm using eigenvectors of the graph Laplacian (EGL-SC) is proposed for image denoising by considering the global structures of images. To exploit the geometry attributes of images, the eigenvectors of the graph Laplacian, which are derived from the graph of noised patches, are incorporated in the sparse model as a set of basis functions. Sequently, the corresponding sparse coding problem is presented and efficiently solved with a relaxed iterative method in the framework of the double sparsity model. Meanwhile, as the denoising performance of the EGL-SC significantly depends on the number of the used eigenvectors, an optimal strategy for the number selection is employed. A parameter called as out-of-control rate is set to record the percentage of the denoised patches that suffer from serious residual errors in the sparse coding procedure. Thus, with the eigenvector number increasing, the appropriate number can be heuristically selected when the out-of-control rate falls below an empirical threshold. Experiments illustrate that the EGL-SC can achieve a better performance than some other well-developed denoising methods, especially in the structural similarity index for the noise of large deviations.

Speech enhancement via sparse coding with ideal binary mask

An improved algorithm is presented for speech enhancement via sparse representation and ideal binary mask (IBM) methods. In the traditional IBM, the basic idea is to identify voiced components as target signal and label unvoiced ones as interference noise vice versa. However, such voiced and unvoiced components still cannot be well separated in target signal and interference noise. To fully exploit the merits of sparse representation theory, we extract the exact voiced component from both the above twofold to obtain the final enhanced speech. Experimental results demonstrate the proposed method can achieve higher PESQ scores than the traditional IBM to efficiently improve speech intelligibility.

Efficient design of sparse FIR filters with optimized filter length

A large number of experiments have demonstrated that for an FIR filter the sparsity of filter coefficients is highly related to its filter order. However, traditional sparse FIR filter design methods focus on how to increase the number of zero-valued coefficients, but overlook the impact of filter orders on design performance. As an attempt to jointly optimize filter length and sparsity of an FIR filter, a novel method is proposed in this paper to design sparse linear-phase FIR filters. With peak error constraints, the objective function of the design problem is formulated as a combination of the sparsity of filter coefficients and a measure of the effective filter order. Then, the design problem is then recast as a weighted l0-norm optimization problem, which is solved by an efficient numerical method based on the iterative-reweighted-least-squares (IRLS) algorithms. Experimental results illustrate that the proposed method can efficiently reduce the effective filter order while enhancing the sparsity of an FIR filter.

Energy efficient Cooperative spectrum sensing with twice selection of nodes

In this paper, we propose a cooperative spectrum sensing method where the nodes detecting the primary user and those sending the sensing results are respectively selected in order to achieve energy efficiency. We formulate the problem of node selection and derive the optimized solution. The results show that the proposed method saves the energy compared to the case where the node selection is not taken into account.

Hybrid framework for image denoising with patch prior estimation

In this paper, a hybrid framework is proposed for image denoising, in which several state-of-the-art denoising methods are efficiently incorporated with a well trade-off by using the prior of patches. In detail, unlike modeling patches with the prior in existed denoising methods, the prior estimation here is presented only to detect the attributes of patches. Then, noisy patches are clustered into several categories according to their patch attributes. Sequentially, different denoising methods are adopted on patches of different categories. The restored image is finally synthesized with the denoised patches of all categories. Experiments show that, by using the hybrid framework, the proposed algorithm is insensitive to the variation of the attributes of images, and can robustly restore images with a remarkable denoising performance.

Image denoising via expected patch log likelihood with Gaussian model identification

In this paper, a new strategy is proposed for the existing expected patch log likelihood (EPLL) algorithm to deal with image denoising, where the Gaussian model identification is incorporated to improve the likelihood estimation for patches. In detail, the noisy patches are first divided into two categories, i.e., smooth and unsmooth groups. Sequentially, in the iteration of the likelihood estimation, these two groups are separately performed via the Gaussian models with the corresponding covariance parameters. Experiments show that, with this Gaussian model identification for patches, the proposed method can achieve the better performance than the traditional EPLL algorithm.

Speech reconstruction via sparse representation using harmonic regularization

In this paper, we consider a speech reconstruction problem, which is efficiently solved via sparse representation. Though a variety of speech reconstruction methods based on the sparse representation are developed, they seldom take into account the intrinsic attributes of speech, e.g., harmonic structures. To address this issue, a harmonic-based sparse representation algorithm is proposed to emphasize harmonic correlations between the adjacent speech frames. Sequently, a corresponding sparse optimal model is presented with a harmonic regularization term, which can be efficiently solved in an iterative framework. Simulation results demonstrate that the proposed method can achieve a better performance to improve the quality of the reconstructed speech than other traditional sparse representation methods.

Blind deblurring with sparse representation via external patch priors

Abstract In this paper, a blind image deblurring method is proposed using sparse representation with external patch priors. Different from traditional sparse-based methods that employ only internal priors from blurred images, additional external information is adopted to reconstruct latent images. In details, the Expected Patch Log Likelihood (EPLL) is introduced as a useful tool to describe external patch priors with a pre-trained Gaussian mixture model. With a set of operations, the EPLL is subsequently incorporated as a regularization term into the existing sparse-based deblurring model. Meanwhile, the dictionary is also carefully designed for each patch of the latent image, where atoms are obtained from the covariance matrix of the corresponding Gaussian component. A deblurring framework is further presented along with our sparse-based model. The solutions are respectively given to efficiently optimize the latent image and the blur kernel with an iterative procedure. The experiments demonstrate that our proposed algorithm achieves a competitive performance compared with the state-of-the-arts. Especially, it not only can obtain more accurate kernels for the deblurring, but also outperforms in noise reduction and artifact suppression for the restored images.

Image enhancement under low luminance with strong light weakening

In low luminance images (night scene), it suffers from various problems, e.g., low overall brightness, poor contrast and serious lack of information. However, in some scenario, strong light also appears in these images. In this case, the regions of high and low luminance both exist, which introduces a more complicated situation for image enhancement. In this paper, we present an enhancement method via dark channel prior to adaptively improve the contrast of given images, especially those containing strong light. To fully use of image dehazing, haze images are firstly obtained from original low luminance images, where the transmittance template is sequentially estimated by the dark channel prior. Later, we design a modified mapping transmittance function to optimize such template, where the factor of strong light is well taken into account. Moreover, to compensate the detail loss in strong light areas, an optimal model is further built to improve the aforementioned template. With a set of dehazing manipulations, enhanced images are finally achieved. Experimental results show that the proposed algorithm not only improves the brightness and contrast of traditional low luminance images, but also can deal with images with strong light, where the regions of strong light are efficiently suppressed.