A smart contract is a computer program which automatically executed with some conditional statements such as "if/then". Since contracts can include vulnerable codes, exploit was recently highlighted one of the severe threats to Ethereum blockchain. As efficient and effective vulnerability detection methods, deep learning methods have been studied due fast speed high accuracy. Recently, using convolutional neural network(CNN) actively classify images transformed from into or invulnerable....

Recently reported research indicate that forest environments have physiological and psychological relaxing effects compared to urban environments. However, some researchers claim the stress of subjects from being watched by others during measurements can affect measurement result in experiments conducted center a street. The present study was determine whether effects, using comparison viewing area with an roof building without others. Near-infrared spectroscopy (NIRS) performed on while...

Basic safety message (BSM) are messages that contain core elements of a vehicle such as vehicle’s size, position, speed, acceleration and others. BSM lightweight can be regularly broadcast by the vehicles to enable variety applications. On other hand, event-driven (EDM) generated at time occurrence accidents or roads sliding much more heavy including pictures, audio videos. Security, architecture communication solutions for use cases have been largely documented on in literature contrary EDM...

GaN epitaxial layers on sapphire substrates were grown by the rotating disk metal organic chemical vapor deposition technique. Excitonic transitions from conduction band to spin-orbit split valence bands observed. At 12 K we observed donor bound exciton and a very weak acceptor exciton. The temperature dependence of luminescence peak positions free-excitons A B fitted Varshni’s equation study variation gap with temperature. linewidth free (A) was studied as function explained theoretical...

Abstract We investigated the efficiency droop and polarization-induced internal electric field of InGaN blue light-emitting diodes (LEDs) grown on silicon(111) c -plane sapphire substrates. The LED sample silicon substrates was considerably lower than that identically fabricated Consequently, showed higher at a sufficiently high injection current despite peak caused by poorer crystal quality. reduced for attributed to its field, which confirmed reverse-bias electro-reflectance measurements...

Abstract To investigate the effects of their surface recovery and optical properties, extremely small sized (12 µm × 12 mesa area) red AlGaInP micro light emitting diodes ( $$\upmu$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mi>μ</mml:mi> </mml:math> LED) were fabricated using a diluted hydrofluoric acid (HF) etch treatment. After chemical treatment, external quantum efficiencies (EQEs) -LED at low high injection current regions have been improved by 35.48% 12.86%,...

As an increasing number of deep-learning-based malware scanners have been proposed, the existing evasion techniques, including code obfuscation and polymorphic malware, are found to be less effective. In this work, we propose a reinforcement learning based semantics-preserving (i.e. functionality-preserving) attack against black-box GNNs (Graph Neural Networks) for detection. The key factor adversarial generation via semantic <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML"...

Magnesium doped GaN epitaxial layers were grown by metal-organic chemical vapor deposition on sapphire substrate. Energy levels of these acceptors investigated systematic photoluminescence measurements in the temperature range 12–300 K. concentration was varied from &amp;lt;1×1019 to higher than 5×1019 cm−3. Photoluminescence made as-grown and annealed samples. We have observed various transitions related donor acceptor their phonon replicas, conduction band acceptors, free excitons. Their...

Abstract This study investigated two artificial intelligence (AI) methods for automatically classifying dental implant diameter and length based on periapical radiographs. The first method, deep learning (DL), involved utilizing the pre-trained VGG16 model adjusting fine-tuning degree to analyze image data obtained from second clustering analysis, was accomplished by analyzing implant-specific feature vector derived three key points coordinates of using k-means++ algorithm weight vector. DL...

An adversarial example, which is an input instance with small, intentional feature perturbations to machine learning models, represents a concrete problem in Artificial intelligence safety. As emerging defense method defend against examples, generative networks-based methods have recently been studied. However, the performance of state-of-the-art limited because target deep neural network models are robust <i>adversarial examples</i> but make false decision for <i>legitimate data</i>. To...

InSb epitaxial layers have been grown on GaAs substrates by low-pressure metalorganic chemical vapor deposition. A 3.15-μm-thick film yielded an x-ray full width at half maximum of 171 arcsec. Hall mobility 76 200 cm2/V s 240 K and a 174 arcsec measured for 4.85-μm-thick epilayer. Measured data shown anomalous behavior. decrease in with decreasing temperature has observed room-temperature increased thickness. In order to explain the data, thickness dependence parameters, coefficient...

We report infrared photoconductors based on InTlSb/InSb grown by low-pressure metalorganic chemical vapor deposition semi-insulating GaAs substrates. The photoresponse spectrum extends up to 8 μm at 77 K. absolute magnitude of the is measured as a function bias. specific detectivity estimated be 3×108 Hz1/2 cm W−1 7 wavelength.

We report the growth of In1−xTlxSb, a new III-V alloy for long-wavelength infrared detector applications, by low-pressure metalorganic chemical vapor deposition. In1−xTlxSb with good surface morphology was obtained on both GaAs and InSb substrates at temperature 455 °C. X-ray diffraction measurements showed resolved peaks films. Infrared absorption spectrum shift toward lower energies compared to spectrum. Hall mobility data In1−xTlxSb/InSb/GaAs structure enhanced low temperatures InSb/GaAs...

InAs0.3Sb0.7 layers with mirrorlike morphology have been grown on GaAs substrates by low-pressure metalorganic chemical vapor deposition. A room-temperature electron Hall mobility of 2×104 cm2/V s has obtained for a 2-μm-thick layer. Low-temperature resistivity the depended TMIn flow rate and layer thickness. decreased monotonically decreasing temperature below 300 K. 77 K conductivity profile shown an anomalous increase in sample thickness except near vicinity heterointerface. In order to...

Differential privacy has emerged as a practical technique for privacy-preserving deep learning. However, recent studies on attacks have demonstrated vulnerabilities in the existing differential implementations models. While encryption-based methods offer robust security, their computational overheads are often prohibitive. To address these challenges, we propose novel privacy-based image generation method. Our approach employs two distinct noise types: one makes unrecognizable to humans,...

We report on high brightness nonpolar a-plane InGaN/GaN LEDs using patterned lateral overgrowth (PLOG) epitaxy. High crystal-quality and smooth surfaces for GaN (a-GaN) films were achieved PLOG with an array of hexagonal SiO2 patterns. The XRC FWHMs as-grown a-GaN found to be 414 317 arcsec (450 455 planar films) along the c-axis m-axis directions, respectively. Plan-view CL clearly reveals periodic patterns higher band edge emission intensity, implying that luminescence properties lying...

Due to the risk of data leakage while training deep learning models in a shared environment, we propose new privacy-preserving (PPDL) method using structural image de-identification approach for object classification. The proposed is designed based on fact that degree distortion an has greatest impact human's perceptual system. Thus, by modifying only parts original one order preserving encryption(OPE), decreases recognition rate human. From experimental results different standard datasets,...

We have measured the temperature and magnetic field dependence of Hall mobility transverse magnetoresistance in n-type InSb films epitaxially grown on GaAs substrates by metalorganic chemical vapor deposition. The show a giant magnetoresistance: e.g., at 240 K resistivity increases over 20 times 5 T; low coefficient 77 is as high 47.5 μΩ cm/G. decreases with saturates higher fields. By taking interface carrier transport into account, observed may be understood based two-layer model.

In the vehicular networks, real-time video reporting service is used to send recorded videos in vehicle cloud. However, when facilitating usage of fourth generation (4G) long term evolution (LTE) was proved suffer from latency while IEEE 802.11p standard does not offer sufficient scalability for a such congested environment. To overcome those drawbacks, fifth-generation (5G)-enabled network considered as promising technology empowering service. this paper, we note that security and privacy...

Adversarial examples are human-imperceptible perturbations to inputs machine learning models. While attacking models, adversarial cause the model make a false positive or negative. So far, two representative defense architectures have shown significant effect: (1) retraining architecture; and (2) input transformation architecture. However, previous methods belonging these do not produce good outputs for every input, i.e., legitimate inputs. Specifically, generate negatives unknown examples,...