Earth Imagery Segmentation on Terrain Surface with Limited Training Labels: A Semi-supervised Approach based on Physics-Guided Graph Co-Training

Given earth imagery with spectral features on a terrain surface, this paper studies surface segmentation based on both explanatory features and surface topology. The problem is important in many spatial and spatiotemporal applications such as flood extent mapping in hydrology. The problem is uniquely challenging for several reasons: first, the size of earth imagery on a terrain surface is often much larger than the input of popular deep convolutional neural networks; second, there exists topological structure dependency between pixel classes on the surface, and such dependency can follow an unknown and non-linear distribution; third, there are often limited training labels. Existing methods for earth imagery segmentation often divide the imagery into patches and consider the elevation as an additional feature channel. These methods do not fully incorporate the spatial topological structural constraint within and across surface patches and thus often show poor results, especially when training labels are limited. Existing methods on semi-supervised and unsupervised learning for earth imagery often focus on learning representation without explicitly incorporating surface topology. In contrast, we propose a novel framework that explicitly models the topological skeleton of a terrain surface with a contour tree from computational topology, which is guided by the physical constraint (e.g., water flow direction on terrains). Our framework consists of two neural networks: a convolutional neural network (CNN) to learn spatial contextual features on a 2D image grid, and a graph neural network (GNN) to learn the statistical distribution of physics-guided spatial topological dependency on the contour tree. The two models are co-trained via variational EM. Evaluations on the real-world flood mapping datasets show that the proposed models outperform baseline methods in classification accuracy, especially when training labels are limited.

Polycrystalline Diamond Coating on Orthopaedic Implants: Realization, and Role of Surface Topology and Chemistry in Adsorption of Proteins and Cell Proliferation

Polycrystalline diamond has the potential to improve the osseointegration of orthopaedic implants compared to conventional osteo-implant materials such as titanium. However, despite the excellent biocompatibility and superior mechanical properties, the major challenge of using diamond for implants such as those used for hip arthroplasty is the limitations of microwave plasma chemical vapor deposition (CVD) techniques to synthesize diamond on complex-shaped objects. Here, for the first time we demonstrate diamond growth on titanium acetabular shells using surface wave plasma CVD method. Polycrystalline diamond coatings were synthesized at low temperatures (~400 °C) on three types of acetabular shells with different surface structure and porosity. We achieved diamond growth on highly porous surfaces designed to mimic the structure of the trabecular bone and improve osseointegration. Biocompatibility was investigated on nanocrystalline diamond (NCD) and ultrananocrystalline diamond (UNCD) coatings terminated either with hydrogen or oxygen. To understand the role of diamond surface topology and chemistry in attachment and proliferation of mammalian cells we investigated adsorption of extracellular matrix (ECM) proteins, and monitored metabolic activity of fibroblasts, osteoblasts, and bone marrow-derived mesenchymal stem cells (MSCs). The interaction of bovine serum albumin (BSA) and Type I collagen with diamond surface was investigated by confocal fluorescence lifetime imaging microscopy (FLIM). We found that proliferation of MSCs was better on hydrogen terminated UNCD than on oxygen terminated counterpart. These findings corelate to the behaviour of collagen on diamond substrates observed by FLIM. Hydrogen terminated UNCD provides better adhesion and proliferation for MSCs, compared to titanium, while growth of fibroblasts is poorest on hydrogen terminated NCD and osteoblasts behave similarly on all tested surfaces. These results open new opportunities for application of diamond coatings on orthopaedic implants.

Влияние химической обработки и топологии поверхности на блокирующее напряжение GaAs тиристорных мезаструктур, выращенных методом ГФЭ МОС

The effect of sulfide passivation (chemical treatment in a peroxide-sulfur etchant and in a solution of Na2S in isopropanol) and complication of the profile of the lateral surface of thyristor mesastructures on the blocking ability of GaAs thyristor mesastructures is investigated. It is shown that the blocking voltage of the chips increases several times both after chemical treatment of the surface and with the complication of the surface topology.

Electrochemically Fabricated Surface-Mesostructured CuNi Bimetallic Catalysts for Hydrogen Production in Alkaline Media

Ni-based bimetallic films with 20 at.% and 45 at.% Cu and mesostructured surfaces were prepared by electrodeposition from an aqueous solution containing micelles of P123 triblock copolymer serving as a structure-directing agent. The pH value of the electrolytic solution had a key effect on both the resulting Cu/Ni ratio and the surface topology. The catalytic activity of the CuNi films toward hydrogen evolution reaction was investigated by cyclic voltammetry (CV) in 1 M KOH electrolyte at room temperature. The Cu45Ni55 film showed the highest activity (even higher than that of a non-mesostructured pure Ni film), which was attributed to the Ni content at the utmost surface, as demonstrated by CV studies, as well as the presence of a highly corrugated surface.

Global perspectives of the bulk electronic structure of URu2Si2 from angle-resolved photoemission

Previous high-resolution angle-resolved photoemission (ARPES) studies of URu2Si2 have characterized the temperature-dependent behavior of narrow-band states close to the Fermi level (E F) at low photon energies near the zone center, with an emphasis on electronic reconstruction due to Brillouin zone folding. A substantial challenge to a proper description is that these states interact with other hole-band states that are generally absent from bulk-sensitive soft x-ray ARPES measurements. Here we provide a more global k-space context for the presence of such states and their relation to the bulk Fermi surface topology using synchrotron-based wide-angle and photon energy-dependent ARPES mapping of the electronic structure using photon energies intermediate between the low-energy regime and the high-energy soft x-ray regime. Small-spot spatial dependence, f-resonant photoemission, Si 2p core-levels, x-ray polarization, surface-dosing modification, and theoretical surface slab calculations are employed to assist identification of bulk versus surface state character of the E F-crossing bands and their relation to specific U- or Si-terminations of the cleaved surface. The bulk Fermi surface topology is critically compared to density functional theory and to dynamical mean field theory calculations. In addition to clarifying some aspects of the previously measured high symmetry Γ, Z and X points, incommensurate 0.6a* nested Fermi-edge states located along Z-N-Z are found to be distinctly different from the density functional theory Fermi surface prediction. The temperature evolution of these states above THO, combined with a more detailed theoretical investigation of this region, suggests a key role of the N-point in the hidden order transition.

Preparation and Characterization of Super-Absorbing Gel Formulated from κ-Carrageenan–Potato Peel Starch Blended Polymers

κ-carrageenan is useful for its superior gelling, hydrogel, and thickening properties. The purpose of the study was to maximize the hydrogel properties and water-absorbing capacity of κ-carrageenan by blending it with starch from potato peels to be used as safe and biodegradable water-absorbent children’s toys. The prepared materials were analyzed using FTIR and Raman spectroscopy to analyze the functional groups. Results showed that there was a shift in the characteristic peaks of starch and κ-carrageenan, which indicated their proper reaction during blend formation. In addition, samples show a peak at 1220 cm−1 corresponding to the ester sulfate groups, and at 1670 cm−1 due to the carbonyl group contained in D-galactose. SEM micrographs showed the presence of rough surface topology after blending the two polymers, with the appearance of small pores. In addition, the presence of surface cracks indicates the biodegradability of the prepared membranes that would result after enzymatic treatment. These results are supported by surface roughness results that show the surface of the κ-carrageenan/starch membranes became rougher after enzymatic treatment. The hydrophilicity of the prepared membranes was evaluated from contact angle (CA) measurements and the swelling ratio. The swelling ratio of the prepared membranes increased gradually as the starch ratio increased, reaching 150%, while the water-uptake capacity increased from 48 ± 4% for plain κ-carrageenan to 150 ± 5% for 1:2 κ-carrageenan/starch blends. The amylase enzyme showed an effective ability to degrade both the plain κ-carrageenan and κ-carrageenan/starch membranes, and release glucose units for up to 236 and 563, respectively. According to these results, these blends could be effectively used in making safe and biodegradable molded toys with superior water-absorbing capabilities.

Dynamometric Investigation on Airborne Particulate Matter (PM) from Friction Materials for Automobile: Impact of Abrasive and Lubricant on PM Emission Factor

Reduction of non-exhaust airborne particulate matter (PM), leading to adverse effects in respiratory system, is an urgent task. In this work, we evaluated the impact of raw materials in friction materials on PM emission due to brake wear for passenger vehicle. Time- and temperature-dependent measurements using dynamometer were made for low-steel friction materials with varied abrasives and lubricant(graphite). The brake emission factor (BEF) for graphite of varied sizes ranged from 6.48 to 7.23 mg/km/vehicle. The number concentration indicates that smaller graphite (10 μm) produces more nano-sized particles than larger size (700 μm) by >50%. Depending on abrasives, BEF was found to be varied as large as by three-times, ranging from 4.37 to 14.41 mg/km/vehicle. As hardness of abrasive increases (SiC > Al2O3 > ZrSiO4), higher BEF was obtained, suggesting that abrasive wear directly contributes to emissions, evidenced by surface topology. Temperature-dependent data imply that particle emission for SiC abrasive is initiated at lower speed in WLTC cycle, where disc temperature (Tdisc) is ~100 °C, than that for ZrSiO4 (Tdisc >120 °C). Analysis of wear debris suggests that larger micron-sized particles include fragmented Fe lumps from disc, whereas smaller particles are, in part, formed by combination of oxidation and aggregation of nano-sized particles into small lumps.