Dai-Gyoung Kim

Bayesian inference and model selection in latent class logit models with parameter constraints: An application to market segmentation

Latent class models have recently drawn considerable attention among many researchers and practitioners as a class of useful tools for capturing heterogeneity across different segments in a target market or population. In this paper, we consider a latent class logit model with parameter constraints and deal with two important issues in the latent class models--parameter estimation and selection of an appropriate number of classes--within a Bayesian framework. A simple Gibbs sampling algorithm is proposed for sample generation from the posterior distribution of unknown parameters. Using the Gibbs output, we propose a method for determining an appropriate number of the latent classes. A real-world marketing example as an application for market segmentation is provided to illustrate the proposed method.

A Fresnelet-Based Encryption of Medical Images using Arnold Transform

Medical images are commonly stored in digital media and transmitted via Internet for certain uses. If a medical information image alters, this can lead to a wrong diagnosis which may create a serious health problem. Moreover, medical images in digital form can easily be modified by wiping off or adding small pieces of information intentionally for certain illegal purposes. Hence, the reliability of medical images is an important criterion in a hospital information system. In this paper, Fresnelet transform is employed along with appropriate handling of the Arnold transform and the discrete cosine transform to provide secure distribution of medical images. This method presents a new data hiding system in which steganography and cryptography are used to prevent unauthorized data access. The experimental results exhibit high imperceptibility for embedded images and significant encryption of information images.

Image de-noising with subband replacement and fusion process using bayes estimators, ☆ ☆☆

Abstract A hybrid image de-noising framework with an automatic parameter selection scheme is proposed to handle substantially high noise with an unknown variance. The impetus of the framework is to preserve the latent detail information of the noisy image while removing the noise with an appropriate smoothing and feasible sharpening. The proposed method is executed in two steps. First, the sub-band replacement and fusion process based on accelerated version of the Bayesian non local means method are implemented to enhance the weak edges that often result in low gradient magnitude and fade out during the de-noising process. Then, a truncated beta-Bernoulli process is employed to infer an appropriate dictionary of the edge enhanced data to obtain de-noising results precisely. Numerical simulations are performed to substantiate the restoration of the weak edges through sub-band replacement and fusion process. The proposed de-noising scheme is validated through visual and quantitative results using well established metrics.

A novel Fresnlet based robust data hiding algorithm for medical images

Robust solutions are recently needed to handle the unusual modification of extensively distributed medical information images. If a medical information image is modified, it may cause to have wrong diagnosis that may arouse a serious problem. Moreover, medical images of digital form are easy to alter for illegal purposes. For instance, by wiping off or adding small information intentionally, it is possible to modify the digital medical images for certain illegal purposes; like insurance claims etc. Hence, the reliability of medical image is an important criterion in a Hospital Information System (HIS). In this paper Fresnelet transform is used for secure distribution of medical images based on Fresnelet coefficient. The experimental results exhibit the high imperceptibility for host image and the significant robustness of embedded image.

Multiscale Hybrid Nonlocal Means Filtering Using Modified Similarity Measure

A new multiscale implementation of nonlocal means filtering (MHNLM) for image denoising is proposed. The proposed algorithm also introduces a modification of the similarity measure for patch comparison. Assuming the patch as an oriented surface, the notion of a normal vectors patch is introduced. The inner product of these normal vectors patches is defined and then used in the weighted Euclidean distance of intensity patches as the weight factor. The algorithm involves two steps: the first step is a multiscale implementation of an accelerated nonlocal means filtering in the discrete stationary wavelet domain to obtain a refined version of the noisy patches for later comparison. The next step is to apply the proposed modification of standard nonlocal means filtering to the noisy image using the reference patches obtained in the first step. These refined patches contain less noise, and consequently the computation of normal vectors and partial derivatives is more precise. Experimental results show equivalent or better performance of the proposed algorithm compared to various state-of-the-art algorithms.

Resolution Enhancement for Digital Off-Axis Hologram Reconstruction

A new method of digital off-axis hologram reconstruction based on the Fresnel transform is proposed. A combination of composite filtering, Abbe’s limitation, and digital lens formulae has been used with an appropriate handling of Fresnel impulse response propagator. A clear image of microscopic object is efficiently reconstructed from hologram using a plane wave with involvement of electric field along bi-cubic interpolation in the final reconstruction step. In particular, the proposed method automatically suppresses the zero order term and virtual image. The image can be reconstructed with large size using interpolation scheme with the Haar wavelet. The proposed method facilitates the transverse high resolution of microscopic image, which has better applicability than other approaches. Moreover, the advantages of this method are its simplicity and convenience in data processing.

Polarization Effects on the Elastic Dust Grain Collisions in Complex Dusty Plasmas

The polarization effects on the elastic dust grain collisions are investigated in unmagnetized complex dusty plasmas. The result shows that the polarization effect enhances the scattering cross section in dusty plasmas. In addition, it is found that the polarization effect on the scattering cross section increases with an increase of the collision velocity. The polarization effect increases with increasing Coulomb radius of interaction between thermal ions and the dust grain in high collision velocities.

Simulation of mask induced polarization effect on imaging in immersion lithography

The minimum feature size of the semiconductor device will be smaller and smaller because of the increasing demand for the high integration of the device. According to recently proposed roadmap, ArF immersion lithography will be used for 65 nm to 45 nm technology nodes. Polarization effect becomes a more important factor due to the increasing demand for high NA optical system and the use of immersion lithography. It is important to know that the polarization effect is induced by mask in small size patterning. The unpolarized plane waves leaving the illumination system are diffracted by the mask. So the light beam going through the mask will experience induced polarization by the mask. In this paper, we considered the change of polarization state as a function of mask properties. We calculated vector diffraction of 193 nm incident light. The masks considered are the chromeless mask, a binary chrome mask and 6 % attenuated phase shifting mask. We use the finite-difference time-domain method to solve the Maxwell equation. The aerial image depends on the polarization states induced by the mask properties such as materials, thickness, and pitch.

Enhanced residual noise estimation of low rank approximation for image denoising

Abstract Recently, the application of rank minimization to image denoising has shown remarkable denoising results. However, due to iterative nature of low rank optimization, the estimation of noise for next iteration is an essential requirement after each iteration. In the existing approaches, the noise is estimated at the level of patches using l2 norm based comparison of the observed image and its denoised version. Because of l2 norm comparison, these patch-wise noise estimates may be affected in the presence of moderate and severe levels of noise. Also, the patch-wise noise estimates are inferred without considering the geometric details of a given image. Moreover, the number of iterations is pre-specified, heuristically, depending on the known initial noise in the image which limits the applicability of the existing algorithms for practical situation where the noise level is not a priori. In this paper, we attempt to address these limitations by introducing the notion of residual noise for iterative rank minimization. The estimation of residual noise also involves image geometry and statistical characteristics along with patch-wise comparison. In addition to more reliable noise estimation, the residual noise can also be employed as an effective stopping criterion for iterative denoising process. Furthermore, the residual noise estimation is a key factor for computation of thresholding weights, which in turn, yield a more reliable truncated singular value decomposition (SVD) for iterative rank minimization algorithms. Based on more precise truncated SVD, we propose to preserve geometric features like edges and texture in the presence of severe noise. Experimental results show that the proposed method provides better or equivalent results as compared to various state-of-the-art denoising algorithms.

The impact of the residual stress on the EUV pellicle

High resolution patterning on the chip could be achieved by extreme ultraviolet lithography (EUVL). However, the defect on the mask becomes more important issue with very short wavelength (13.5 nm). Using the pellicle which could protect the mask from the defects can support high volume manufacturing (HVM). Most of the materials considered for pellicle have relatively high extinction coefficient in EUV region. Therefore, the thickness of the pellicle should be ~ nm thin. The stress of the pellicle is dependent not only on the temperature but also on the mechanical properties of the pellicle. The stress induced by the gravity was small compared to the thermal stress. However, the residual stress should be also considered since it is dependent on the pellicle manufacturing environment and this stress is comparable with the thermal stress. Our result shows the importance of the lowering the pellicle fabrication temperature in terms of the extending the lifetime during the scanning process.