Uniform Region Research Articles

A Survey on Deep Learning Techniques for Stereo-Based Depth Estimation.

Estimating depth from RGB images is a long-standing ill-posed problem, which has been explored for decades by the computer vision, graphics, and machine learning communities. Among the existing techniques, stereo matching remains one of the most widely used in the literature due to its strong connection to the human binocular system. Traditionally, stereo-based depth estimation has been addressed through matching hand-crafted features across multiple images. Despite the extensive amount of research, these traditional techniques still suffer in the presence of highly textured areas, large uniform regions, and occlusions. Motivated by their growing success in solving various 2D and 3D vision problems, deep learning for stereo-based depth estimation has attracted a growing interest from the community, with more than 150 papers published in this area between 2014 and 2019. This new generation of methods has demonstrated a significant leap in performance, enabling applications such as autonomous driving and augmented reality. In this paper, we provide a comprehensive survey of this new and continuously growing field of research, summarize the most commonly used pipelines, and discuss their benefits and limitations. In retrospect of what has been achieved so far, we also conjecture what the future may hold for deep learning-based stereo for depth estimation research.

Influence of Cryogenic Treatment on Microstructure and Mechanical Properties of Ti6Al4V Alloy

Cryogenic treatment is an emerging application that can make significant changes in many materials. In this study, shallow (− 80 °C) and deep (− 196 °C) cryogenic treatment was applied to Ti6Al4V alloy. Additionally, aging treatment and supplementary deep cryogenic treatment were also performed. Tensile tests were done to observe the effects of cryogenic treatment on the mechanical properties and formability of Ti6Al4 alloy. According to the results of these tests, it was observed that the cryogenic treatment had a positive influence on the plasticity properties of Ti6Al4V alloy. For the 24-h and 36-h deep cryo-treated samples, the uniform plastic deformation region (UPDR) was extended 5 and 8.3% compared to untreated sample, respectively. However, this improvement in plasticity observed in samples treated with cryogenic treatment for 36 h led to a decrease in yield strength by 2%. Microstructural and XRD analysis was performed to characterize the microstructural changes that happened with the effect of applied heat treatments. The characterization study shows that with the application of cryogenic treatment internal structure of the material changed, the β phase ratio of the alloy decreased from 8.1 to 5.6% with 36-h deep cryogenic treatment. The unstable β phase was transformed into α phase and stable β phase. The experimental study shows that cryogenic treatment improves the plasticity and toughness of Ti6Al4V alloy at room temperature and regulates the microstructure and reduces the residual stress.

Oxidation behavior analysis of a ferritic ODS steel in supercritical Water

Abstract The corrosion behavior of 14Cr ODS steel exposed to the supercritical water (SCW) at 600 °C under 25 MPa with the dissolved oxygen content of 300 ppb was investigated. The weight gain measurement and corrosion layers analysis were used to determine the corrosion behavior and corrosion mechanism of the 14Cr ODS steel. The results indicate that the 14Cr ODS steel exposed in this work possesses better corrosion resistance than the other ODS steel with a similar composition. Both uniform corrosion regions and nodular corrosion regions can be seen on the exposed specimens, which is due to the bimodal microstructure of the ODS steel. The corrosion products have a multi-layer structure, including the outer oxide layer, the inner oxide layer, and the transition layer. The oxide in the inner layer is FeCr2O4. The oxides in the outer layer change from the initial Fe3O4 to Fe2O3 with the increase of exposure time. The possible corrosion mechanism is analyzed as well, combining with the microstructure characters of the 14Cr ODS steel.

Embedding in medical images with contrast enhancement and tamper detection capability

This paper presents an efficient data hiding technique capable of providing improved visual quality of watermarked images, besides having the ability to detect the tamper, if any. It is a spatial domain approach in which major emphasis is on improving the visual quality rather than increasing the PSNR or the embedding capacity. The medical images have been divided into Region of Interest (ROI) and Non-Region of Interest (NROI). Bringing out details that lie within the low dynamic range is very important in medical images for effective diagnosis. ROI being diagnostically critical region is enhanced using contrast stretching and subsequently, data is reversibly embedded into the peak bins of ROI. Only those peak bins are employed for reversible data embedding that have an adjacent empty bin to overcome the problem of overflow and underflow. In NROI, the uniform intensity and redundant information region, Least Significant Bit (LSB) embedding is employed for increasing the payload. For tamper detection, a fragile watermark has been embedded in the ROI. To evaluate the scheme various parameters like peak signal to noise ratio (PSNR), No-Reference Quality Metric for contrast-distorted images (NR-CDIQA) and Structural Similarity Index Matrix (SSIM) have been calculated. The experimental results show a remarkable increase in visual quality compared to state-of-art.

Stress–Strain Properties of Artificially Aged 6061 Al Alloy: Experiments and Modeling

The paper examines the tensile deformation behavior of the Al-Mg-Si alloy (6061 Al alloy) subjected to various aging conditions. 6061 Al alloy is commonly used in aerospace/aircraft industry due to its performance on corrosion resistance, formability, and weldability. Five different heat treatment procedures, including T4 natural aging and T6 peak-strength temper conditions, were designed to investigate the effect of artificial aging on the mechanical behavior of the alloy. Tensile tests were performed to determine the stress–strain behavior of the material both in uniform and non-uniform deformation regions. Mechanical material properties including yield, ultimate and fracture strengths, uniform and total strains, hardness, strength coefficient, and strain-hardening exponent were obtained experimentally. The relationship between equivalent strain, equivalent stress, and hardness is also examined. The fracture strength of specimens was determined to be less than the Holloman model predictions for the fracture strains of specimens. Void development, which is dependent on the amount of plastic strain development, is determined to be the main reason for this discrepancy between the Holloman model and fracture stress. To calculate the homogeneous stress in the metal matrix of the porous domain, Eshelby-based Mori–Tanaka method (MTM) was used. The calculated average stress in the metal matrix shows good agreement with the Holloman equation predictions. Thus, void development explains the interrupted strain hardening after necking.

Effect of interphase parameters on elastic modulus prediction for cellulose nanocrystal fiber reinforced polymer composite

Abstract In this study, the effective elastic modulus (Ec ) of a cellulose nanocrystal fiber reinforced polymer composite was evaluated using the Mori-Tanaka and finite element (FE) model. The FE model was generated using a representative volume element with a periodic boundary condition. The mass fractions of the fiber in the composites (MFf ) were set to 1, 2, and 3 wt.%. Elastic modulus values for interphase were input and were either uniform or exhibited a gradient. The E c for the uniform interphase region increased significantly with MFf , but was relatively low for interphase regions exhibiting a gradient. The results show that interphase parameters must be considered carefully when predicting Ec using the FE model.

A practical octree liquid simulator with adaptive surface resolution

We propose a new adaptive liquid simulation framework that achieves highly detailed behavior with reduced implementation complexity. Prior work has shown that spatially adaptive grids are efficient for simulating large-scale liquid scenarios, but in order to enable adaptivity along the liquid surface these methods require either expensive boundary-conforming (re-)meshing or elaborate treatments for second order accurate interface conditions. This complexity greatly increases the difficulty of implementation and maintainability, potentially making it infeasible for practitioners. We therefore present new algorithms for adaptive simulation that are comparatively easy to implement yet efficiently yield high quality results. First, we develop a novel staggered octree Poisson discretization for free surfaces that is second order in pressure and gives smooth surface motions even across octree T-junctions, without a power/Voronoi diagram construction. We augment this discretization with an adaptivity-compatible surface tension force that likewise supports T-junctions. Second, we propose a moving least squares strategy for level set and velocity interpolation that requires minimal knowledge of the local tree structure while blending near-seamlessly with standard trilinear interpolation in uniform regions. Finally, to maximally exploit the flexibility of our new surface-adaptive solver, we propose several novel extensions to sizing function design that enhance its effectiveness and flexibility. We perform a range of rigorous numerical experiments to evaluate the reliability and limitations of our method, as well as demonstrating it on several complex high-resolution liquid animation scenarios.

Clinical evaluation of three respiratory gating schemes for different respiratory patterns on cardiac SPECT.

Respiratory gating reduces respiratory blur in cardiac single photon emission computed tomography (SPECT). It can be implemented as three gating schemes: (a) equal amplitude-based gating (AG); (b) phase or time-based gating (TG); or (c) equal count-based gating (CG), that is, a variant of amplitude-based method. The goal of this study is to evaluate the effectiveness of these respiratory gating methods for patients with different respiratory patterns in myocardial perfusion SPECT. We reviewed 1274 anonymized patient respiratory traces obtained via the Vicon motion-tracking system during their 99m Tc-sestamibi SPECT scans and grouped them into four breathing categories: (a) regular respiration (RR); (b) periodic respiration (PR); (c) respiration with apnea (AR); and (d) unclassified respiration (UR). For each respiratory pattern, 15 patients were randomly selected and their list-mode data were rebinned using the three gating schemes. A preliminary reconstruction was performed for each gate with the heart region segmented and registered to a reference gate to estimate the respiratory motion. A final reconstruction incorporating respiratory motion correction was done to get a final image set. The estimated respiratory motion, the full-width-at-half-maxima (FWHM) measured across the image intensity profile of the left ventricle wall, as well as the normalized standard deviation measured in a uniform cuboid region of the thorax were analyzed. There are 47.1%, 24.3%, 13.5%, and 15.1% RR, PR, AR, and UR patients, respectively, among the 1274 patients in this study. The differences among the three gating schemes in RR were smaller than other respiratory patterns. The AG and CG methods showed statistically larger motion estimation than TG particularly in the AR and PR patterns. Noise of AG varied more in different gates, especially for AR and UR patterns. More than half of the patients reviewed exhibited nonregular breathing patterns. Amplitude-based gating, that is, AG and CG, is a preferred gating method for such patterns and is a robust respiratory gating implementation method given the respiratory pattern of the patients is unknown before data acquisition. Phase gating is also a feasible option for regular respiratory pattern.

CO2 laser-based side-polishing of silica optical fibers.

In this Letter, an optical fiber side-polishing process is proposed that is non-contact, versatile, and scalable. A CO2 laser, with carefully selected pulse parameters, is used to remove cladding material from the side of an optical fiber in a controlled manner. The resulting side-polished optical fiber has adiabatic polishing transitions and a flat uniform polished region. The technique provides a pristine polishing surface with an RMS surface roughness of less than 2 nm. Furthermore, in contrast to traditional side-polishing methods, the wear of hard tooling, the associated surface flaws, and issues with residual abrasive particulates are all negated. It is anticipated that this technique will provide a robust platform for the next generation of optical fiber devices that are based on in-fiber light-matter interaction with exotic materials, such as low-dimensional semi-conductors and topological insulators.

Electron density and effective atomic number estimation in a maximum a posteriori framework for dual-energy computed tomography.

The stoichiometric calibration method for dual-energy CT (DECT) proposed by Bourque et al. (Phys Med Biol. 59:2059; 2014), which provides estimators of the electron density and the effective atomic number, is adapted to a maximum a posteriori (MAP) framework to increase the model's robustness to noise and biases in CT data, specifically for human tissues. Robust physical parameter estimation from noisy DECT scans is required to maximize the precision of quantities used for radiotherapy treatment planning such as the proton stopping power (SPR). Estimation of electron density and effective atomic number is performed by constraining their variation to the natural range of values expected for human tissues, while maximizing attenuation data fidelity. The MAP framework is first compared against the original method using theoretical CT numbers with Gaussian noise. The quantitative accuracy of the MAP framework is then validated experimentally on the Gammex 467 phantom. Then, using two clinical datasets, the advantages of the approach are experimentally evaluated, qualitatively, and quantitatively. The theoretical study shows that the root-mean-square error on the electron density, the effective atomic number and the SPR are, respectively, reduced from 2.3 to 1.5, 5.7 to 3.2 and 2.8 to 1.7% with the adapted framework, when analyzing soft tissues and bone together. The experimental validation study shows that the standard deviation in Gammex inserts can be reduced, on average, by factors of 1.4 (electron density), 2.7 (effective atomic number), and 1.9 (SPR), while the quantitative accuracy of the three physical parameters is preserved, on average. Evaluation on clinical datasets show apparent noise reduction in maps of all estimated physical quantities, and suggests that the MAP framework has increased robustness to beam hardening and photon starvation artifacts. Mean values for the electron density, the effective atomic number, and the SPR averaged in four uniform regions of interest (brain, muscle, adipose, and cranium), respectively, differ by 0.7, 1.8, and 0.9% between both frameworks. The standard deviation in the same regions of interest is also reduced, on average, by factors of 1.8, 6.6, and 3.2 with the MAP framework. Differences in mean value and standard deviations are statistically significant. Theoretical and experimental results suggest that the MAP framework produces more accurate and precise estimates of the electron density and SPR. Thus, the present approach limits the propagation of noise in DECT attenuation data to radiotherapy-related parameters maps such as the SPR and the electron density. Using a MAP framework with DECT for radiotherapy treatment planning can help maximizing the precision of dose calculation. The method also provides more precise estimates of the effective atomic number. The MAP methodology is presented in a general way such that it can be adapted to any DECT image-based tissue characterization method.

Noise and spatial resolution properties of a commercially available deep learning-based CT reconstruction algorithm.

To characterize the noise and spatial resolution properties of a commercially available deep learning-based computed tomography (CT) reconstruction algorithm. Two phantom experiments were performed. The first used a multisized image quality phantom (Mercury v3.0, Duke University) imaged at five radiation dose levels (CTDIvol : 0.9, 1.2, 3.6, 7.0, and 22.3mGy) with a fixed tube current technique on a commercial CT scanner (GE Revolution CT). Images were reconstructed with conventional (FBP), iterative (GE ASiR-V), and deep learning-based (GE True Fidelity) reconstruction algorithms. Noise power spectrum (NPS), high-contrast (air-polyethylene interface), and intermediate-contrast (water-polyethylene interface) task transfer functions (TTF) were measured for each dose level and phantom size and summarized in terms of average noise frequency (fav ) and frequency at which the TTF was reduced to 50% (f50% ), respectively. The second experiment used a custom phantom with low-contrast rods and lung texture sections for the assessment of low-contrast TTF and noise spatial distribution. The phantom was imaged at five dose levels (CTDIvol : 1.0, 2.1, 3.0, 6.0, and 10.0mGy) with 20 repeated scans at each dose, and images reconstructed with the same reconstruction algorithms. The local noise stationarity was assessed by generating spatial noise maps from the ensemble of repeated images and computing a noise inhomogeneity index, , following AAPM TG233 methods. All measurements were compared among the algorithms. Compared to FBP, noise magnitude was reduced on average (± one standard deviation) by 74±6% and 68±4% for ASiR-V (at "100%" setting) and True Fidelity (at "High" setting), respectively. The noise texture from ASiR-V had substantially lower noise frequency content with 55±4% lower NPS fav compared to FBP while True Fidelity had only marginally different noise frequency content with 9±5% lower NPS fav compared to FBP. Both ASiR-V and True Fidelity demonstrated locally nonstationary noise in a lung texture background at all radiation dose levels, with higher noise near high-contrast edges of vessels and lower noise in uniform regions. At the 1.0mGy dose level values were 314% and 271% higher in ASiR-V and True Fidelity compared to FBP, respectively. High-contrast spatial resolution was similar between all algorithms for all dose levels and phantom sizes (<3% difference in TTF f50% ). Compared to FBP, low-contrast spatial resolution was lower for ASiR-V and True Fidelity with a reduction of TTF f50% of up to 42% and 36%, respectively. The deep learning-based CT reconstruction demonstrated a strong noise magnitude reduction compared to FBP while maintaining similar noise texture and high-contrast spatial resolution. However, the algorithm resulted in images with a locally nonstationary noise in lung textured backgrounds and had somewhat degraded low-contrast spatial resolution similar to what has been observed in currently available iterative reconstruction techniques.

Development of a Broad Bandwidth Helmholtz Coil for Biomagnetic Application

The alternating uniform magnetic field is useful for the biomagnetic study with magnetic nanoparticles. However, for the coil system driven by the magnetic field generator with a conventional low-voltage power source, it is challenging to achieve the desired field strength with a large coil due to the large impedance at higher frequencies. This article proposed a Helmholtz coil system that can generate the dc and ac magnetic field up to 1.3 kHz with low distortion. In order to overcome the inductive load, a computer-controlled parallel capacitor array was designed and built to tune the series coil-capacitor circuit to the resonant frequency. Experimental results show that the coil system has a bandwidth over 1 kHz with a 1% uniform region covering the 10 × 10 mm sample area. The proposed system is promising for the application in biomagnetic assay using the tunnel magnetoresistance sensors.

Investigation of the flow characteristics with one-line emergent canopy patches in open channel

Experiments are performed on channels with emergent one-line porous canopy patches distributed along the channel sidewalls in which flow dynamics and turbulent structures are analyzed. The evolution of the transverse distribution of the mean streamwise velocity and Reynolds stress profiles in specific transverse sections with downstream distance suggests that a fully equilibrated zone is reached beyond an adjustment distance, and the adjustment distance needed for vegetation with high solid volume fractions is short. The flow field can be divided into three main different regions based on the transverse distribution of the mean streamwise velocity and Reynolds stress in the equilibrium zone: a boundary-affected region, an analog mixing layer, and a uniform region in a non-vegetated area. The analog mixing layer exhibits a two-layer structure governed by local coherent vortices. The dominant frequencies of the vortices at the interface are determined through spectral analysis and proven to correspond to a patch scale. The development of the shear layer, the analysis of the frequencies and scales of the vortices, and the experiment on flow visualization indicate that von-Karman vortices continue to form downstream despite the interference of the sidewall and the special one-line distribution pattern of patches. Quadrant analysis is adopted to illustrate the lateral momentum exchange caused by different motions. Results show that ejections and sweeps are the main contributions to Reynolds stress in the analog mixing layer. Different signs of Reynolds stress in the boundary-affected region and the analog mixing layer can also be explained through quadrant analysis.

Mapping small and medium-sized water reservoirs using Sentinel-1A: a case study in Chiapas, Mexico

Using satellite data to study small water bodies (SWB) and medium-sized water bodies (MSWB) is extremely useful for understanding their status, how to conserve them as water reservoirs, and their vulnerability to climate variability. The images studied in our work correspond to different-sized lagoons located in areas with high and low topography in a tropical region of Chiapas, Mexico. Our research project delineates SWB and MSWB. For this analysis, we considered water bodies to be uniform regions in a synthetic aperture radar image. The robustness of the method was determined based on an analysis of the morphologies of 23 lagoons. Several methods, including Hermite transform, were analyzed and compared with other image denoising methods used to improve speckle reduction. To obtain additional spatial information for image classification, we analyzed texture using the gray-level co-occurrence matrix. The results indicate that the Hermite filter is the best method for identifying water bodies. The advantage of this filter is the identification of local patterns such as edges and lines. It also preserves and improves aspects related to the homogeneity of water bodies, using the Hermite coefficient selection criteria for local pattern feature selection/extraction. The lake water extent products demonstrate that Sentinel-1 is useful for identifying SWB in this study area. The results show very high detection of water bodies, with adequate detection for water bodies larger than 2 ha, and an area accuracy of 80%.

Effects of Plastic Strain on Mechanical and Electrical Conductivity Response of (100) Monocrystalline Copper

The physical characteristics of material would be influenced by its stress states. In this paper, nanoindentation experiments with a maximum load of 100 mN and strain-sensitive resistance tests were conducted on (100) monocrystalline copper with tension-induced plastic strain to investigate the influences of plastic strain on mechanical and electrical characteristics. By theoretical and experimental analysis, indentation depth, elastic modulus, recovery rate of total work, contact area, pile-up and apparent hardness of the deformed and virgin material were obtained and the corresponding investigation was compared and analyzed. The experimental results revealed that under the same conditions, tensile pre-deformed material would like to generate more penetration depth to achieve the same load during indentation. As the plastic strain increased in the uniform plastic region, both the total and elastic indentation work increased, while the recovery rate of total work, contact area, apparent hardness, and elastic modulus decreased. Meanwhile, the contribution of plastic strain to electrical resistivity was also investigated. The quantitative relation between electrical resistivity and plastic strains was extracted from experiments. The understanding of the role of plastic strain in mechanical and electrical properties of monocrystalline copper would be helpful to the application in damage detection systems and strain sensing sensor.

Otwarty dostęp w krajach globalnego Południa

Cel/Teza: Artykuł referuje bieżący etap rozwoju otwartego dostępu w krajach globalnego Południa, ze wskazaniem podobieństw i odmienności pomiędzy różnymi obszarami świata, nakreśleniem specyfiki reprezentatywnych krajów lub grupy krajów oraz określeniem przyczyn ograniczenia ich aktywności w tym zakresie.Koncepcja/Metody badań: Termin „globalne Południe”, odnoszący się do mniej rozwiniętych krajów Azji, Pacyfiku, Afryki i Ameryki Łacińskiej, stał się ostatnio popularny w dyskursie akademickim, chociaż nie został prawidłowo zdefiniowany. W artykule różne przypadki jego użycia są analizowane w konfrontacji z terminem „globalna Północ”. Analizie poddane zostały również dane dotyczące otwartego dostępu w krajach lobalnego Południa dostępne w źródłach internetowych.Wyniki i wnioski: Poszczególne obszary globalnego Południa są pod wieloma względami zróżnicowane, choć wszystkie łączy niższy status ekonomiczny i zespół rozmaitych problemów z tego wynikających. Granice pomiędzy Północą i Południem zmieniają się, w pewnych miejscach zaś zacierają. Środowiska naukowe w tych krajach różnią się pod względem potrzeb, dążeń i wyobrażeń na temat tego, jak powinno, ich zdaniem, wyglądać publikowanie prac naukowych. Ze względu na owe specyficzne cechy próba dopasowania narzędzi wypracowanych w naukowych ośrodkach globalnej Północy może okazać się szkodliwa. Kluczowa okazuje się komplementarność informacji (jej złożoność, organizacja i weryfikacja), która, dzięki narzędziom technicznym, staje się dostępna szerszej naukowej społeczności.Oryginalność/Wartość poznawcza: W polskiej literaturze badawczej zagadnienie to pojawia się rzadko. Artykuł zbiera i zestawia dane dotyczące wszystkich obszarów globalnego Południa obecne głównie w literaturze anglojęzycznej.

Hardening effect of multi-energyW2+-ion irradiation on tungsten–potassium alloy**Project supported by the National Natural Science Foundation of China (Grant Nos. 11975160 and 11775149). One of the authors, Kun Zhang, was supported by the Fundamental Research Funds for the Central Universities, China.

Tungsten is one of the most promising plasma-facing materials (PFMs) to be used in the nuclear fusion reactor as divertor material in the future. In this work, W2+-ions bombardment is used to simulate the neutron irradiation damage to commercial pure tungsten (W) and rolled tungsten–potassium (W–K). The 7 MeV of 3 × 1015 W2+-ions/cm2, 3 MeV of 4.5 × 1014 W2+, and 2 MeV of 3 × 1014 W2+-ions/cm2 are applied at 923 K in sequence to produce a uniform region of 100 nm–400 nm beneath the sample surface with the maximum damage value of 11.5 dpa. Nanoindentation is used to inspect the changes in hardness and elastic modulus after self-ion irradiation. Irradiation hardening occurred in both materials. The irradiation hardening of rolled W–K is affected by two factors: one is the absorption of vacancies and interstitial atoms by potassium bubbles, and the other is the interaction between potassium bubbles and dislocations. Under the condition of 11.5 dpa, the capability of defect absorption can reach a threshold. As a result, dislocations finally dominate the hardening of rolled W–K. Specific features of dislocation loops in W–K are further observed by transmission electron microscopy (TEM) to explain the hardening effect. This work might provide valuable enlightenment for W–K alloy as a promising plasma facing material candidate.

Attentive feature integration network for detecting salient objects in images

Benefiting from the development of convolutional neural networks, salient object detection has yielded a qualitative leap in performance. In recent years, most of deep learning based methods utilize multi-level features and obtain inferred saliency map in a coarse-to-fine manner. However, how to learn and represent powerful features is still a challenge. In this paper, we propose a novel FCN-like approach named attentive feature integration network (AFINet) for pixel-wise salient object detection, which results saliency maps with explicit boundary and uniform highlighted regions. Specifically, it adopts feature enhancement module (FEM) to extract rich and enhanced features from backbone net. A feature discrimination module (FDM) is designed to utilize the predicted saliency map generated by deeper layer to help shallower layer learn useful and discriminative attentive features. Moreover, we introduce the saliency information from deeper layer to the shallower one in saliency prediction module (SPM), which helps shallow side outputs accurately locate salient regions. In addition, we design a saliency fusion module (SFM) to integrate different side outputs for utilizing multi-level features. Finally, a fully connected CRF scheme can be optimally incorporated for obtaining saliency results with a higher accuracy. Both qualitative and quantitative comparisons and evaluations conducted on five publicly benchmark datasets demonstrate that our proposed approach compares favorably against 17 state-of-the-art approaches.

Attention and boundary guided salient object detection

In recent years, fully convolutional neural network (FCN) has broken all records in various vision task. It also achieves great performance in salient object detection. However, most of the state-of-the-art methods have suffered from the challenge of precisely segmenting the entire salient object with uniform region and explicit boundary and effectively suppressing the backgrounds on complex images. There is still a large room for improvement over the FCN-based saliency detection approaches. In this paper, we propose an attention and boundary guided deep neural network for salient object detection to better locate and segment the salient objects with uniform interior and explicit boundary. A channel-wise attention module is utilized to emphasize the important regions, which selects the important feature channels and assigns large weights to them. A boundary information localization module is proposed for suppressing the irrelevant boundary information to better locate and explore the useful structure of objects. The proposed approach achieves state-of-the-art performance on four well-known benchmark datasets.

Design of Lu2O3-reinforced Cf/SiC-ZrB2-ZrC ultra-high temperature ceramic matrix composites: Wetting and interfacial reactivity by ZrSi2 based alloys

The wettability and infiltration of molten ZrSi2 and ZrSi2-Lu2O3 alloys into Cf/SiC and B4C-infiltrated Cf/SiC composites were investigated to understand the interfacial interactions that occur during the development of Cf/SiC-ZrC and Cf/SiC-ZrB2-ZrC-Lu2O3 materials. A significant evaporation of Si from the liquid affected the wetting behaviour of the alloy when tested in a vacuum at 1670 °C. The better wetting and spreading of the alloy over the surface was observed for the composites with lower overall porosity (12 %). On the other hand, the formation of an outer dense layer, followed up by the uniform infiltrated region up to ∼ 1 mm was observed for the Cf/SiC with higher porosity (21 %). The infiltrated alloy reacted with SiC matrix to form ZrC or with B4C-infiltrated SiC matrix to form ZrB2-ZrC-SiC. The Lu2O3 particles were not wetted by the melt, and were pushed away of the reaction zone by the solidification front.