We study theoretically the propagation of electromagnetic waves through periodic structures consistent layered materials with an intensity-dependent dielectric constant. find transmission properties to be strongly modulated by both frequency and intensity in presence nonlinearity. The diagram versus amplitude plane exhibits distinctive features depending upon whether Kerr coefficient is positive or negative. These features, though complicated, can understood analysis stable orbits...

The anisotropic tight-binding model, which is defined by a disordered Hamiltonian with transfer-energy-matrix element in the z direction, t, different than one x-y plane studied using finite-size scaling methods. dependence of mobility edge on strength anisotropy t (0\ensuremath{\le}t\ensuremath{\le}1) obtained for center band, E=0. Even very low-t values an appreciable amount disorder needed to localize E=0 state. These results are found be satisfactory agreement predictions potential-well...

We have studied numerically the soliton scattering from one or two impurities in one-dimensional atomic lattices with nonlinear first-neighbor interactions of both quartic and Morse type. Our results for small-amplitude incident solitons are interpreted some analytical formulas obtained linear theories, while comparison wave-packet made low- high-amplitude solitons. Finally, interference effects studied.

In addition to the inherent qualities that biometrics posses, powerful signal processing tools enabled widespread deployment of biometric-based identification/verification systems. However, due nature biometric data, well-established cryptographic (such as hashing, encryption, etc.) are not sufficient for solving one most important problems related systems, namely, template security. this paper, we examine and show how apply a recently proposed secure sketch scheme in order protect...

Fake faces produced by deepfake techniques have attracted public concerns in recent years. Deepfake detection is a binary classification task that distinguishes fake from real ones. As the training data for usually generated via various face forgery methods, it difficult single decision boundary to distinguish faces. Besides that, learned features often involve irrelevant information identifying are diverse methods. To deal with such challenges, unlike existing approaches regard as...

The scattering of a soliton from disordered one-dimensional atomic lattice with nonlinear nearest-neighbor interactions quartic type is studied numerically. disorder the binary-alloy concentration impurity masses $m$ given by $p$. We numerically find that for large enough lengths $L$, transmission coefficient $T$ decays as $\frac{1}{\sqrt{L}}$. This behavior has been obtained also an analytical study Gaussian wave packet in linear system. For short and intermediate lengths, different power...

A zero-knowledge watermark detector allows an owner to prove a verifier that image in question indeed contains the owner's without revealing much information about actual watermark. In such scenario, publishes committed before detection so as show she knows detection. However, this does not imply can work appeared public. One well known counter example is invertibility/ambiguity attacks where adversary create ambiguous situation by deriving forged from published work, and commits...

Deepfake detection automatically recognizes the manipulated media by analyzing whether it contains forgeries generated through deep learning. It is natural to ask which among existing deepfake approaches stand out as top performers. This question pivotal for identifying promising research directions and offering practical guidance. Unfortunately, conducting a sound benchmark comparison of popular based on literature results challenging due inconsistent evaluation conditions across studies....

Channel feature detectors have shown great advantages in human detection. However, a large pool of channel features extracted for detection usually contains many redundant and irrelevant features. To address this issue, we propose robust discriminative weighted sparse partial least square approach selection apply it to Unlike squares (PLS), which is straightforward dimensionality reduction technique, using PLS achieve selection. Furthermore, order obtain latent matrix, formulate regularized...

Channel feature detectors are the most popular approaches for pedestrian detection recently. However, of these train boosted decision trees by selecting a single at each node, which does not effectively exploit multi-feature cues and spatial information. To address this issue, paper proposes to construct co-occurrence multiple channel features in local image neighborhoods detection. In our approach, binary pattern is represented combining variables quantized from feature, information...

We use the correlated-random-field approximation and a conditional coherent-potential to study ground-state phase diagram of one-band Hubbard model. Our approach incorporates both spin electron correlations. It has been found that for certain values parameters model, in addition paramagnetic, ferromagnetic, antiferromagnetic, spin-glass or spin-liquid phases our earlier studies, phase-separated is be favored 0.7\ensuremath{\le}n\ensuremath{\le}1.0 U/B\ensuremath{\ge}2 good agreement with...

Although it is clear that security an important issue in digital watermarking applications,the main concerns addressed by the current literature are robustness, capacity and imperceptibility. The inadequacy of prevailing design paradigm tackling issues mainly due to incomplete assessment threat model. goal this paper take a detailed rigorous look at model for variety applications. In extended abstract, we outline requirements few common applications explore more detail specific application...

In augmented reality systems, the three dimensional registration technology based on artificial markers has been widely used. The algorithm is divided into two steps: (1) marker detection and recognition, (2) estimation of camera pose. this paper, a novel designed projective invariants principle, such as collinearity between points cross-ratio. Compared with used in ARToolkit, our new can greatly reduce storage space templates. And marker, recognition proposed. By using algorithm, AR system...

The success of clustering algorithms in object segmentation depends on the quality evaluation criterion. However, sonar images are seriously affected by noise. Most existing criteria such as Davies Bouldin (DB) criterion only considers their pixel value, and image information extraction is not sufficient. As a result, they fail to achieve good underwater results. To overcome this problem, study proposes an improved DB with pulse ‐coupled neural network (PCNN), which called DB‐PCNN. In...

Scaling of the conductances and finite-size localization lengths is generalized to anisotropic systems tested in two dimensional systems. functions isotropic are recovered once dimension system each direction chosen be proportional length. It also shown that geometric mean a function conductivities. The ratio square root conductivities, which turn anisotropy strength t, weak scattering limit.

The integrated densities of states (DOS) one-, two-, and three-dimensional tight-binding models with diagonal off-diagonal disorder are calculated numerically. Our results for the diagonal-disorder case confirm expected universal behavior DOS, particularly near band edges. simple exponential DOS is clearly displayed in results. off-diagonal-disorder follows a too, which agrees reasonably well coherent-potential-approximation (CPA) compared various phenomenological theories, CPA conclusions,...

The captured images and videos in bad weather usually have degraded quality by reduced contrast faded colors, is very difficult to achieve promising performance. traditional prior techniques are not sufficient address this challenging problem deal with the halo artifacts brightness distortion problems. In paper, we propose a multi-scale fusion method for single fog image restoration. By creating two divided regions, global atmospheric light can be effectively obtained sky regions. To...

In many applications, it is often desirable to extract a consistent key from multimedia object (e.g., an image), even when the has gone through noisy channel. For example, extracted can be used generate content dependent watermarks mitigate copy attacks, or for two more parties establish session their versions of same object. Robust hash functions are useful in extracting such keys. It differs cryptographic that small noise messages would yield value with high probability. However, security...

<abstract> <p>Limited by GPU memory, high-resolution image segmentation is a highly challenging task. To improve the accuracy of segmentation, High-Resolution Refine Net (HRRNet) proposed. The network divided into rough module and refinement module. former improves DeepLabV3+ adding shallow features 1/2 original size corresponding skip connection to obtain better results, output which used as input latter. In module, first, global context information obtained process. Second,...

In this paper, we design and realize an augmented reality interaction framework. A linear interpolation algorithm in adaptive spatial grid is proposed to calibrate the magnetic tracker, calibrated data used for registration of user's viewpoint virtual objects. An optimized collision detection based on Vclip adopted. Additionally, grasp gesture recognition rough sets theory presented capture recognize hand operation. order make it easier learn new Chinese characters, a character teaching...