Portrait Segmentation Using Ensemble of Heterogeneous Deep-Learning Models

Image segmentation plays a central role in a broad range of applications, such as medical image analysis, autonomous vehicles, video surveillance and augmented reality. Portrait segmentation, which is a subset of semantic image segmentation, is widely used as a preprocessing step in multiple applications such as security systems, entertainment applications, video conferences, etc. A substantial amount of deep learning-based portrait segmentation approaches have been developed, since the performance and accuracy of semantic image segmentation have improved significantly due to the recent introduction of deep learning technology. However, these approaches are limited to a single portrait segmentation model. In this paper, we propose a novel approach using an ensemble method by combining multiple heterogeneous deep-learning based portrait segmentation models to improve the segmentation performance. The Two-Models ensemble and Three-Models ensemble, using a simple soft voting method and weighted soft voting method, were experimented. Intersection over Union (IoU) metric, IoU standard deviation and false prediction rate were used to evaluate the performance. Cost efficiency was calculated to analyze the efficiency of segmentation. The experiment results show that the proposed ensemble approach can perform with higher accuracy and lower errors than single deep-learning-based portrait segmentation models. The results also show that the ensemble of deep-learning models typically increases the use of memory and computing power, although it also shows that the ensemble of deep-learning models can perform more efficiently than a single model with higher accuracy using less memory and less computing power.     

Conformational preferences and pKa value of cysteine residue

The conformational preferences of the Cys dipeptides with thiol and thiolate groups (Ac-Cys-NHMe and Ac-Cys (-)-NHMe, respectively) and the apparent (i.e., macroscopic) p K a value of the Cys dipeptide have been studied at the hybrid density functional B3LYP/6-311++G(d,p)//B3LYP/6-31+G(d) level with the conductor-like polarizable continuum model in the gas phase and in water. The hydrogen bonds and/or favorable interactions between the backbone and the thiol group of the side chain resulted in the different conformational preferences of the Cys and Cys (-) dipeptides from those of the Ala dipeptide in the gas phase and in water, although the preferred conformations of the Cys dipeptide are in part similar to those of the Ala dipeptide. In particular, the interactions between the thiolate group and the backbone amide groups appear to play a role in stabilizing the alpha- or 3 10-helical conformations for the Cys (-) dipeptide in the gas phase and in water. The p K a value of the Cys residue is estimated to be 8.58 at 25 degrees C using the statistically weighted free energies of all feasible conformations for the Cys and Cys (-) dipeptides in the gas phase and solvation free energies, which is consistent with the observed values of 8.3 and 8.22 +/- 0.16.     

State-selective frustration as a key driver of allosteric pluripotency

Allosteric pluripotency arises when an allosteric effector switches from agonist to antagonist depending on the experimental conditions. For example, the Rp-cAMPS ligand of Protein Kinase A (PKA) switches from agonist to antagonist as the MgATP concentration increases and/or the kinase substrate affinity or concentration decreases. Understanding allosteric pluripotency is essential to design effective allosteric therapeutics with minimal side effects. Allosteric pluripotency of PKA arises from divergent allosteric responses of two homologous tandem cAMP-binding domains, resulting in a free energy landscape for the Rp-cAMPS-bound PKA regulatory subunit R1a in which the ground state is kinase inhibition-incompetent and the kinase inhibition-competent state is excited. The magnitude of the free energy difference between the ground non-inhibitory and excited inhibitory states (ΔG_R,Gap) relative to the effective free energy of R1a binding to the catalytic subunit of PKA (ΔG_R:C) dictates whether the antagonism-to-agonism switch occurs. However, the key drivers of ΔG_R,Gap are not fully understood. Here, by analyzing an R1a mutant that selectively silences allosteric pluripotency, we show that a major determinant of ΔG_R,Gap unexpectedly arises from state-selective frustration in the ground inhibition-incompetent state of Rp-cAMPS-bound R1a. Such frustration is caused by steric clashes between the phosphate-binding cassette and the helices preceding the lid, which interact with the phosphate and base of Rp-cAMPS, respectively. These clashes are absent in the excited inhibitory state, thus reducing the ΔG_R,Gap to values comparable to ΔG_R:C, as needed for allosteric pluripotency to occur. The resulting model of allosteric pluripotency is anticipated to assist the design of effective allosteric modulators.

The Rp-cAMPS ligand of protein kinase A switches from agonist to antagonist depending on metabolite and proteomic contexts. We show that the state-selective frustration is a key driver of this allosteric pluripotency phenomenon.     

W 1, p(·)-Regularity for Elliptic Equations with Measurable Coefficients in Nonsmooth Domains

We establish global W ^{1, p(·)}-estimates for second order elliptic equations in divergence form under the natural assumption that p(·) is log-Hölder continuous. To this end, we assume that the coefficients are measurable in one variable and have small BMO semi-norms in the other variables and the boundary of the domain is Reifenberg flat. Our work is an optimal and natural extension of W ^1,p -regularity for such equations with merely measurable coefficients beyond Lipschitz domains.     

Effect of gamma irradiation on Burkholderia thailandensis (Burkholderia pseudomallei surrogate) survival under combinations of pH and NaCl

This study evaluated the effect of gamma irradiation on Burkholderia thailandensis (Burkholderia pseudomallei surrogate; potential bioterrorism agent) survival under different levels of NaCl and pH. B. thailandensis in Luria Bertani broth supplemented with NaCl (0–3%), and pH-adjusted to 4–7 was treated with gamma irradiation (0–0.5 kGy). Surviving cell counts of bacteria were then enumerated on tryptic soy agar. Data for the cell counts were also used to calculate D₁₀ values (the dose required to reduce 1 log CFU/mL of B. thailandensis). Cell counts of B. thailandensis were decreased (P<0.05) as irradiation dose increased, and no differences (P≥0.05) in cell counts of the bacteria were observed among different levels of NaCl and pH. D₁₀ values ranged from 0.04 to 0.07 kGy, regardless of NaCl and pH level. These results indicate that low doses of gamma irradiation should be a useful treatment in decreasing the potential bioterrorism bacteria, which may possibly infect humans through foods.     

Synthesis and characterization of chemically modified polystyrene as processable carbon fiber precursors

In this study, the chemically modified polystyrene was studied for improvement of polystyrene which had low carbon yield, as a carbon fiber precursor. The polystyrene was synthesized with divinylbenzene which was used as a cross-linking agent by the solution polymerization method. Then the synthesized polystyrene was nitrated with sulfuric and nitric acids solution (H₂SO₄/HNO₃) followed by reduction to form nitrogen-functional groups. The surface properties of the modified polystyrene were investigated by Fourier transfer infrared spectroscopy to confirm the introduction of functional groups on the modified polystyrene surfaces. The thermal properties of the modified polystyrene were measured by thermogravimetric analysis and differential scanning calorimetry. The morphologies of the electrospun polystyrene fibers by a chemical modification were observed by scanning electron microscopy. From the results, the nitrogen-functional groups were introduced on the modified polystyrene surfaces which affected the quantity of functional groups. Also, the chemical treatment affected the carbon yield of the polystyrene owing to the introduction of nitrogen-functional groups on polystyrene surfaces. Consequently, it was concluded that the chemical treatment of polystyrene fibers enables it to be used as a possible carbon fiber precursor.     

Polymer-based electrospray device with multiple nozzles to minimize end effect phenomenon

In this paper, “end effect”, which represents the deflection of meniscus of boundary nozzles in an array due to asymmetric electric field and repulsive forces between adjacent nozzles, is experimentally investigated and a polymer-based electrospray device is fabricated and tested to minimize the end effect. For the electrically conductive nozzles of electrospray, the end effect could be reduced by dummy nozzles, through which no liquid is supplied, mounted at the boundary of the array. However, the polymer-based electrospray device can eliminate the end effect without using the dummy dry nozzles due to dielectric characteristic. A novel multiple nozzles of Polymethyl Methacrylate (PMMA)-based electrospray device was successfully fabricated and evaluated for reducing the end effect, showing no deflection of menisci through the boundary nozzles. And ten nozzles device was fabricated and observed to eject liquid jet simultaneously. The electrical current of the spray issued from the nozzles of the polymer-based electrospray device is measured with respect to flow rate and shows a good agreement with the scaling law.     

Cysteine-179 of IkappaB kinase beta plays a critical role in enzyme activation by promoting phosphorylation of activation loop serines

IkappaB kinase beta (IKKbeta) subunit of IKK complex is essential for the activation of NF-kappaB in response to various proinflammatory signals. Cys-179 in the activation loop of IKKbeta is known to be the target site for IKK inhibitors such as cyclopentenone prostaglandins, arsenite, and antirheumatic gold compounds. Here we show that a mutant IKKbeta in which Cys-179 is substituted with alanine had decreased activity when it was expressed in HEK-293 cells, and TNF stimulation did not restore the activity. Phosphorylation of activation loop serines (Ser-177 and Ser-181) which is required for IKKbeta activation was reduced in the IKKbeta (C179A) mutant. The activity of IKKbeta (C179A) was partially recovered when its phosphorylation was enforced by coexpression with mitogen-activated protein kinase kinase kinases (MAPKKK) such as NF-kappaB inducing kinase (NIK) and MAPK/extracellular signal-regulated kinase kinase kinase 1(MEKK1) or when the serine residues were replaced with phospho-mimetic glutamate. The IKKbeta (C179A) mutant was normal in dimer formation, while its activity abnormally responded to the change in the concentration of substrate ATP in reaction mixture. Our results suggest that Cys-179 of IKKbeta plays a critical role in enzyme activation by promoting phosphorylation of activation-loop serines and interaction with ATP.     

A simple taste analyzing system design for visual and quantitative analysis of different tastes using multi-array chemical sensors and pattern recognition techniques

In this paper, we describe design of a simple taste analyzing system using sensory system based on a multi-array chemical sensor (MACS) and personal digital assistant (PDA) for visual and quantitative analysis of different tastes using pattern recognition techniques. The sensory system is communicated with PDA, which has several interesting benefits for data analysis and display, via wireless using the Bluetooth. A various pattern recognition techniques are adapted including spider map, principal component analysis (PCA) and fuzzy C-means (FCM) algorithm to classify visually data patterns detected by the sensory system. The proposed techniques can be determined the cluster centers and membership grade of patterns through the unsupervised way. The membership grade of an unknown pattern, which does not shown previously, can be visually and analytically determined. Throughout the experimental trails, the taste analyzing system is demonstrated robust performance through data acquisition via wireless communication and visual and quantitative analysis of different tastes for the liquids. The system, which is implemented as a simple hand-held taste analyzing instrument, can be applicable to on-site taste monitoring.     

Ambient xenon sampling using an Ag/ZSM-5 zeolite

Journal of Radioanalytical and Nuclear Chemistry - An ambient xenon sampling system using Ag/ZSM-5 was designed, and its performance was tested. The xenon collection efficiency of the system was...