Comparing U-Net Based Models for Denoising Color Images

Digital images often become corrupted by undesirable noise during the process of acquisition, compression, storage, and transmission. Although the kinds of digital noise are varied, current denoising studies focus on denoising only a single and specific kind of noise using a devoted deep-learning model. Lack of generalization is a major limitation of these models. They cannot be extended to filter image noises other than those for which they are designed. This study deals with the design and training of a generalized deep learning denoising model that can remove five different kinds of noise from any digital image: Gaussian noise, salt-and-pepper noise, clipped whites, clipped blacks, and camera shake. The denoising model is constructed on the standard segmentation U-Net architecture and has three variants—U-Net with Group Normalization, Residual U-Net, and Dense U-Net. The combination of adversarial and L1 norm loss function re-produces sharply denoised images and show performance improvement over the standard U-Net, Denoising Convolutional Neural Network (DnCNN), and Wide Interface Network (WIN5RB) denoising models.

Assessment of Gold Bio-Functionalization for Wide-Interface Biosensing Platforms

The continuous improvement of the technical potential of bioelectronic devices for biosensing applications will provide clinicians with a reliable tool for biomarker quantification down to the single molecule. Eventually, physicians will be able to identify the very moment at which the illness state begins, with a terrific impact on the quality of life along with a reduction of health care expenses. However, in clinical practice, to gather enough information to formulate a diagnosis, multiple biomarkers are normally quantified from the same biological sample simultaneously. Therefore, it is critically important to translate lab-based bioelectronic devices based on electrolyte gated thin-film transistor technology into a cost-effective portable multiplexing array prototype. In this perspective, the assessment of cost-effective manufacturability represents a crucial step, with specific regard to the optimization of the bio-functionalization protocol of the transistor gate module. Hence, we have assessed, using surface plasmon resonance technique, a sustainable and reliable cost-effective process to successfully bio-functionalize a gold surface, suitable as gate electrode for wide-field bioelectronic sensors. The bio-functionalization process herein investigated allows to reduce the biorecognition element concentration to one-tenth, drastically impacting the manufacturing costs while retaining high analytical performance.

Quantifying Disorder in Networks

The authors introduce a novel entropic notion with the purpose of quantifying disorder/uncertainty in networks. This is based on the Laplacian and it is exactly the von Neumann entropy of certain quantum mechanical states. It is remarkable that the von Neumann entropy depends on spectral properties and it can be computed efficiently. The analytical results described here and the numerical computations lead us to conclude that the von Neumann entropy increases under edge addition, increases with the regularity properties of the network and with the number of its connected components. The notion opens the perspective of a wide interface between quantum information theory and the study of complex networks at the statistical level.

Quantifying Complexity in Networks

The authors introduce a novel entropic notion with the purpose of quantifying disorder/uncertainty in networks. This is based on the Laplacian and it is exactly the von Neumann entropy of certain quantum mechanical states. It is remarkable that the von Neumann entropy depends on spectral properties and it can be computed efficiently. The analytical results described here and the numerical computations lead us to conclude that the von Neumann entropy increases under edge addition, increases with the regularity properties of the network and with the number of its connected components. The notion opens the perspective of a wide interface between quantum information theory and the study of complex networks at the statistical level.

Characteristics of Alumina Diffusion Barrier Films on Hastelloy

The diffusion behavior of elements constituting Hastelloy C-276 (C, Si, Mn, Co, W, Fe, Cr, Mo, and Ni) in alumina films was investigated using secondary ion mass spectroscopy. The films were deposited by ion-beam-assisted deposition and annealed in vacuum over a temperature range of 500–1000 °C. Characterization of film microstructure was performed using transmission electron microscopy and selected area diffraction analyses. The films were predominantly amorphous with alumina nanocrystallites nonuniformly dispersed throughout the volume both before and after annealing. A relatively wide interface region between the Hastelloy substrate and alumina film was formed in the as-deposited sample due to ion beam mixing. No diffusion of any of the substrate elements was observed after annealing, except for Mn, Cr, and Ni. The impurity depth distributions consisted of two components, which differed by several orders of magnitude with respect to diffusion coefficient and solubility. Activation energies and temperature dependencies of the diffusion coefficients were determined, and a diffusion mechanism was discussed.

Bone-Prosthesis Implant Relation in Total Cementless Hip Arthroprosthesis

The best relation between bone and prosthetic implant occurs when an adequate bone surface preparation and a correct implant placing ensure wide interface contact, with homogeneous load distribution (primary stability). This is necessary premise to achieve osteointegration (secondary stability) which is conditioned by factors connected with the implant (materials, surface work, lining) and with the hosting bone (pathological onset, sex, age). The evolution of bone-implant relation leads to the formation of fibrous or osseous interface; the stimulus which act on the interface (in relation with primary stability, distance between surfaces, bony heritage, etc…) infact determine a fibroblastic or osteoblastic evolution of mesenchymal elements colonizing the interface during the initial phases. In successive phases the prevailing apposition (hyperthropy) or resorption (focal or diffuse atrophy) processes is equally conditioned by mechanical factors (the position and dimension of the implant).

A Turbulent Drier-Extractor and a Wide-Interface Extractor

A simpler modification of an original turbulent drier-extractor with provision for measuring moisture is presented, as well as methods for determination of lipids in serum and in urine in the wide-interface extractor, and a new use for the wide-interface extractor in washing extracts from the drier-extractor. The two extraction units described may also serve as a reference standard and may be preferred for routine lipid assays.