Video-based isolated hand sign language recognition using a deep cascaded model

In this paper, we propose an efficient cascaded model for sign language recognition taking benefit from spatio-temporal hand-based information using deep learning approaches, especially Single Shot Detector (SSD), Convolutional Neural Network (CNN), and Long Short Term Memory (LSTM), from videos. Our simple yet efficient and accurate model includes two main parts: hand detection and sign recognition. Three types of spatial features, including hand features, Extra Spatial Hand Relation (ESHR) features, and Hand Pose (HP) features, have been fused in the model to feed to LSTM for temporal features extraction. We train SSD model for hand detection using some videos collected from five online sign dictionaries. Our model is evaluated on our proposed dataset (Rastgoo et al., Expert Syst Appl 150: 113336, 2020), including 10’000 sign videos for 100 Persian sign using 10 contributors in 10 different backgrounds, and isoGD dataset. Using the 5-fold cross-validation method, our model outperforms state-of-the-art alternatives in sign language recognition

     

Ag–Sr doped mesoporous bioactive glass nanoparticles loaded chitosan/gelatin coating for orthopedic implants

In this study, we investigated surface and biological properties of Ag–Sr-doped mesoporous bioactive glass nanoparticle (Ag–Sr MBGN) loaded chitosan/gelatin coatings deposited by electrophoretic deposition (EPD) on 316L stainless steel. The EPD parameters, that is, deposition time, applied voltage, and distance between the electrodes was optimized by the Taguchi design of experiment (DoE) approach. Scanning electron microscopy (SEM) images illustrated the spherical morphology of the synthesized Ag–Sr MBGNs with the mean particle size of 160 ± 20 nm. Energy-dispersive X-ray (EDX) spectroscopy results confirmed the presence of Ag and Sr in the synthesized MBGNs. Optimum EPD parameters determined by DoE approach were 5 g/L of Ag–Sr MBGNs, deposition time of 5 min, and applied voltage of 30 V. SEM images confirmed that the coatings were fairly homogenous. Fourier-transform infrared spectroscopy and EDX results confirmed the presence of chitosan, gelatin, and Ag–Sr MBGNs in the coatings. Chitosan/gelatin/Ag–Sr MBGN composite coatings exhibited suitable wettability for the protein attachment and proliferation of osteoblast cells. The composite coatings exhibited suitable adhesion strength with the substrate. The coatings developed HA crystals upon immersion in simulated body fluid. The results of the turbidity test confirmed that the coatings are antibacterial to the Escherichia coli cells.     

Synthesis of 3D nanostructured metal alloy of immiscible materials induced by megahertz-repetition femtosecond laser pulses

In this work, we have proposed a concept for the generation of three-dimensional (3D) nanostructured metal alloys of immiscible materials induced by megahertz-frequency ultrafast laser pulses. A mixture of two microparticle materials (aluminum and nickel oxide) and nickel oxide microparticles coated onto an aluminum foil have been used in this study. After laser irradiation, three different types of nanostructure composites have been observed: aluminum embedded in nickel nuclei, agglomerated chain of aluminum and nickel nanoparticles, and finally, aluminum nanoparticles grown on nickel microparticles. In comparison with current nanofabrication methods which are used only for one-dimensional nanofabrication, this technique enables us to fabricate 3D nanostructured metal alloys of two or more nanoparticle materials with varied composite concentrations under various predetermined conditions. This technique can lead to promising solutions for the fabrication of 3D nanostructured metal alloys in applications such as fuel-cell energy generation and development of custom-designed, functionally graded biomaterials and biocomposites.     

Machine learning based optimized live virtual machine migration over WAN links

Live virtual machine migration is one of the most promising features of data center virtualization technology. Numerous strategies have been proposed for live migration of virtual machines on local area networks. These strategies work perfectly in their respective domains with negligible downtime. However, these techniques are not suitable to handle live migration over wide area networks and results in significant downtime. In this paper we have proposed a Machine Learning based Downtime Optimization (MLDO) approach which is an adaptive live migration approach based on predictive mechanisms that reduces downtime during live migration over wide area networks for standard workloads. The main contribution of our work is to employ machine learning methods to reduce downtime. Machine learning methods are also used to introduce automated learning into the predictive model and adaptive threshold levels. We compare our proposed approach with existing strategies in terms of downtime observed during the migration process and have observed improvements in downtime of up to 15 %.     

Synergistic effect of nanoparticles and Thalassospira sp. isolated from Bahregan area in biodegradation of oil sludge

Oil sludge is obtained from the contaminated site of Bahregan area in the Persian Gulf in Iran. Chemotaxis was used for the isolation of alkane-degrading bacteria from oil sludge and the alk genes were determined with specific primers. bac1 identified in Thalassospira was selected as a powerful strain for biodegradation of oil sludge. Biodegradation of oil sludge by bac1 in the presence of nanoparticles was investigated by GC-MS analysis. Synergistic effect of Fe₂O₃ and Fe₃O₄ nanoparticles on bacteria increases the biodegradation of oil sludge and produces cyclosiloxane compound which is used in the field of medicine; however, the Thalassospira reduced its toxicity in the environment.     

Revisiting the free transverse vibration of embedded single-layer graphene sheets acted upon by an in-plane magnetic field

Murmu et al. [23] recently presented a nonlocal model for the transverse vibration of simply supported graphene sheets in the presence of a unidirectional in-plane magnetic field. Further studies showed that the majority of Lorentz’s force components were improperly provided and led to invalid governing equations. To remove such deficiencies, the most general form of Lorentz’s force components is carefully extracted in the present work. The nonlocal equations of motion of the problem are reconstructed and solved again. The influences of crucial parameters on the flexural frequencies of magnetically affected graphene sheets and nanoribbons are examined in detail. Furthermore, the crucial discrepancies between the results obtained in this study and those of the abovementioned previous work are rationally discussed. Some erroneous results of the latter are also rectified.     

A GDQ approach to thermally nonlinear generalized thermoelasticity of disks

Generalized thermoelasticity response of an annular disk subjected to thermal shock on its inner surface is analyzed in this research. The Lord–Shulman theory, which accounts for one relaxation time in the conventional Fourier law, is used to avoid the infinite speed of thermal wave propagation. Unlike the other available works in which the first law of thermodynamics is linearized, the nonlinearity arising from the temperature change is taken into consideration. The first law of thermodynamics in this case becomes nonlinear and the analysis under such formulation is called thermally nonlinear. Two coupled equations, i.e., the radial displacement wave equation and temperature wave propagation equation are obtained. These equations and the associated boundary conditions are discreted through the generalized differential quadrature method. Solution of the time-dependent system of equations is obtained using the Newmark time marching scheme and the successive Picard method. Numerical results are provided for both thermally linear and thermally nonlinear temperature and radial displacement wave propagations. Parametric studies reveal that at higher temperature levels, thermally nonlinear first law of thermodynamics should be considered instead of thermally linear one. Furthermore, the higher the coupling parameter and/or relaxation time, the higher the divergence of thermally nonlinear-/linear-based results.     

Highly efficient synthesis of benzopyranopyridines via ZrP2O7 nanoparticles catalyzed multicomponent reactions of salicylaldehydes with malononitrile and thiols

ZrP₂O₇ nanoparticles as an efficient catalyst have been used for the preparation of benzopyrano[2,3-b]pyridines from the four-component condensation reaction of salicylalde-hydes, thiols, and 2 equiv. of malononitrile under reflux conditions in ethanol in excellent yields and short reaction times.     

On the Erosion–Corrosion Behavior of Active Screen Plasma Nitrided St52 Steel

In this work, samples of St52 steel were plasma nitrided using an iron screen, in an N₂: H₂ gas mixture ratio of 4 : 1, at 500 and 550°C for 5, 10 and 15 h. The X-ray diffraction and optical microscopy methods were used for structural characterization of the coatings. Results indicated that the coatings were composed of Fe_2–3N and Fe₄N phases growing at longer deposition times. Moreover, the Fe_2–3N phase was decomposed to Fe₄N after 10 h of plasma nitriding. The erosion–corrosion behavior of nitrided coatings and a bare substrate were studied in various impact angles: 30, 45, 60 and 90 degrees. Polarization curves of the coated and uncoated samples were recorded between–900 to 600 mV, as a function of the slurry impact angle. Results showed that an active screen plasma nitriding method significantly enhanced the erosion–corrosion resistance of the St52 steel. Moreover, an impact angle of 30° on the sample surface yielded a lower weightloss whereas increasing the impact angle up to 90° caused more weight-loss due to the brittle characteristic of iron nitride coatings. According to SEM micrographs, by increasing the impact angle up to 90°, the depth of the removed mass increased substantially.     

Nonlinear free vibration of temperature-dependent sandwich beams with carbon nanotube-reinforced face sheets

In this research, large amplitude free vibrations of a sandwich beam with stiff core and carbonnanotube (CNT)-reinforced face sheets are analysed. The distribution of CNTs across the thickness of the facesheets may be uniform or functionally graded. The equivalent single- layer theory of Timoshenko is used toconstruct the Hamiltonian of the beam under the von Kármán type of geometrical nonlinearity assumptions.A uniform temperature field through the beam is also included in the formulation. The Ritz method withpolynomial basis functions is used to discrete the equations of motion and establish the matrix representation ofthe governing equations. A nonlinear eigenvalue problem is obtained and solved using a standard continuationprocedure. After validating the developed solution method and formulation, parametric studies are conductedto examine the influences of thermal environment, core thickness-to-face sheet thickness ratio, boundaryconditions, amplitude of vibrations, CNTs volume fraction and their distribution pattern. It is concludedthat an increase in the volume fraction of CNTs results in higher fundamental frequency and decreases thenonlinear-to-linear frequency ratio.