Informatica

Aimed at achieving the accurate restoration of Poissonian images that exhibit neat edges and no staircase effect, this article develops a novel hybrid nonconvex double regularizer model. The proposed scheme closely takes the advantages of total variation with overlapping group sparsity and nonconvex high-order total variation priors. The overlapping group sparsity is adopted to globally suppress the staircase artifacts, while the nonconvex high-order regularization plays the role of locally preserving the significant image features and edge details. Computationally, a quite efficient alternating direction method of multipliers, associated with the iteratively reweighted ${\ell _{1}}$ algorithm and the majorization-minimization method, is employed to settle the optimization problem iteratively. Finally, exhaustive simulation experiments are executed for recovering Poissonian images, which are made comparisons with several state-of-the-art restoration strategies, indicate the brilliant performance of our model in terms of intuitive effects and accuracy evaluation.

A method for counterfactual explanation of machine learning survival models is proposed. One of the difficulties of solving the counterfactual explanation problem is that the classes of examples are implicitly defined through outcomes of a machine learning survival model in the form of survival functions. A condition that establishes the difference between survival functions of the original example and the counterfactual is introduced. This condition is based on using a distance between mean times to event. It is shown that the counterfactual explanation problem can be reduced to a standard convex optimization problem with linear constraints when the explained black-box model is the Cox model. For other black-box models, it is proposed to apply the well-known Particle Swarm Optimization algorithm. Numerical experiments with real and synthetic data demonstrate the proposed method.

Medical X-ray images are prevalent and are the least expensive diagnostic imaging method available widely. The handling of film processing and digitization introduces noise in X-ray images and suppressing such noise is an important step in medical image analysis. In this work, we use an adaptive total variation regularization method for removing quantum noise from X-ray images. By utilizing an edge indicator measure along with the well-known edge preserving total variation regularization, we obtain noise removal without losing salient features. Experimental results on different X-ray images indicate the promise of our approach. Synthetic examples are given to compare the performance of our scheme with traditional total variation and anisotropic diffusion methods from the literature. Overall, our proposed approach obtains better results in terms of visual appearance as well as with respect to different error metrics and structural similarity.

The choice of natural image prior decides the quality of restored image. Recently successful algorithms exploit heavy-tailed gradient distribution as image prior to restore latent image with piecewise smooth regions. However, these prior assumed also remove the mid-frequency component such as textural details regions while they preserve sharp edges. That because gradient profile in fractal-like texture do not have sparse characteristic.

In this paper we have proposed a novel method for image denoising using local polynomial approximation (LPA) combined with the relative intersection of confidence intervals (RICI) rule. The algorithm performs separable column-wise and row-wise image denoising (i.e., independently by rows and by columns), combining the obtained results into the final image estimate. The newly developed method performs competitively among recently published state-of-the-art denoising methods in terms of the peak signal-to-noise ratio (PSNR), even outperforming them for small to medium noise variances for images that are piecewise constant along their rows and columns.