A Light-Weight Deep Learning-Based Architecture for Sign Language Classification

With advancements in computing powers and the overall quality of images captured on everyday cameras,a much wider range of possibilities has opened in various scenarios.This fact has several implications for deaf and dumb people as they have a chance to communicate with a greater number of people much easier.More than ever before,there is a plethora of info about sign language usage in the real world.Sign languages,and by extension the datasets available,are of two forms,isolated sign language and continuous sign language.The main difference between the two types is that in isolated sign language,the hand signs cover individual letters of the alphabet.In continuous sign language,entire words’hand signs are used.This paper will explore a novel deep learning architecture that will use recently published large pre-trained image models to quickly and accurately recognize the alphabets in the American Sign Language(ASL).The study will focus on isolated sign language to demonstrate that it is possible to achieve a high level of classification accuracy on the data,thereby showing that interpreters can be implemented in the real world.The newly proposed Mobile-NetV2 architecture serves as the backbone of this study.It is designed to run on end devices like mobile phones and infer signals(what does it infer)from images in a relatively short amount of time.With the proposed architecture in this paper,the classification accuracy of 98.77%in the Indian Sign Language(ISL)and American Sign Language(ASL)is achieved,outperforming the existing state-of-the-art systems.

IC Pattern Based Power Factor Maximization Model for Improved Power Stabilization

The voltagefluctuation in electric circuits has been identified as key issue in different electric systems.As the usage of electricity growing in rapid way,there exist higherfluctuations in powerflow.To maintain theflow or stabi-lity of power in any electric circuit,there are many circuit models are discussed in literature.However,they suffer to maintain the output voltage and not capable of maintaining power stability.To improve the performance in power stabilization,an efficient IC pattern based power factor maximization model(ICPFMM)in this article.The model is focused on improving the power stability with the use of IC(Inductor and Conductor)towards identifying most efficient circuit for the current duty cycle according to the input voltage,voltage in capacitor and output voltage required.The model with boost converter diverts the incoming voltage through number of conductors and inductors.By triggering specific inductor,a specific capacitor gets charged and a particular circuit gets on.The model maintains num-ber of IC(Inductor and Conductor)patterns through which the powerflow occurs.According to that,the pattern available,the mofset controls the level of power to be regulated through any circuit.From the pattern,the model computes the Cir-cuits Switching Loss and Circuits Conduction Loss for various circuits.Accord-ing to the input voltage,the model estimates Circuit Power Stabilization Support(CPSS)according to the voltage available in any capacitor and input voltage.Using the value of CPSS,the model trigger optimal number of circuits to maintain voltage stability.In this approach,more than one circuit has been triggered to maintain output voltage and to get charged.The proposed model not only main-tains power stability but also reduces the wastage in voltage which is not utilized.The proposed model improves the performance in voltage stability with less switching loss.

Speech Separation Methodology for Hearing Aid

In the design of hearing aids(HA),the real-time speech-enhancement is done.The digital hearing aids should provide high signal-to-noise ratio,gain improvement and should eliminate feedback.In generic hearing aids the perfor-mance towards different frequencies varies and non uniform.Existing noise can-cellation and speech separation methods drops the voice magnitude under the noise environment.The performance of the HA for frequency response is non uni-form.Existing noise suppression methods reduce the required signal strength also.So,the performance of uniform sub band analysis is poor when hearing aid is con-cern.In this paper,a speech separation method using Non-negative Matrix Fac-torization(NMF)algorithm is proposed for wavelet decomposition.The Proposed non-uniformfilter-bank was validated by parameters like band power,Signal-to-noise ratio(SNR),Mean Square Error(MSE),Signal to Noise and Dis-tortion Ratio(SINAD),Spurious-free dynamic range(SFDR),error and time.The speech recordings before and after separation was evaluated for quality using objective speech quality measures International Telecommunication Union-Telecommunication standard ITU-T P.862.

IM-EDRD from Retinal Fundus Images Using Multi-Level Classification Techniques

In recent years,there has been a significant increase in the number of people suffering from eye illnesses,which should be treated as soon as possible in order to avoid blindness.Retinal Fundus images are employed for this purpose,as well as for analysing eye abnormalities and diagnosing eye illnesses.Exudates can be recognised as bright lesions in fundus pictures,which can be thefirst indicator of diabetic retinopathy.With that in mind,the purpose of this work is to create an Integrated Model for Exudate and Diabetic Retinopathy Diagnosis(IM-EDRD)with multi-level classifications.The model uses Support Vector Machine(SVM)-based classification to separate normal and abnormal fundus images at thefirst level.The input pictures for SVM are pre-processed with Green Channel Extraction and the retrieved features are based on Gray Level Co-occurrence Matrix(GLCM).Furthermore,the presence of Exudate and Diabetic Retinopathy(DR)in fundus images is detected using the Adaptive Neuro Fuzzy Inference System(ANFIS)classifier at the second level of classification.Exudate detection,blood vessel extraction,and Optic Disc(OD)detection are all processed to achieve suitable results.Furthermore,the second level processing comprises Morphological Component Analysis(MCA)based image enhancement and object segmentation processes,as well as feature extraction for training the ANFIS classifier,to reliably diagnose DR.Furthermore,thefindings reveal that the proposed model surpasses existing models in terms of accuracy,time efficiency,and precision rate with the lowest possible error rate.

Deep Learning-Based FOPID Controller for Cascaded DC-DC Converters

Smart grids and their technologies transform the traditional electric grids to assure safe,secure,cost-effective,and reliable power transmission.Non-linear phenomena in power systems,such as voltage collapse and oscillatory phenomena,can be investigated by chaos theory.Recently,renewable energy resources,such as wind turbines,and solar photovoltaic(PV)arrays,have been widely used for electric power generation.The design of the controller for the direct Current(DC)converter in a PV system is performed based on the linearized model at an appropriate operating point.However,these operating points are everchanging in a PV system,and the design of the controller is usually accomplished based on a low irradiance level.This study designs a fractional-order proportional-integrated-derivative(FOPID)controller using deep learning(DL)with quasi-oppositional Archimedes Optimization algorithm(FOPID-QOAOA)for cascaded DC-DC converters in micro-grid applications.The presented FOPIDQOAOA model is designed to enhance the overall efficiency of the cascaded DC-DC boost converter.In addition,the proposed model develops a FOPID controller using a stacked sparse autoencoder(SSAE)model to regulate the converter output voltage.To tune the hyper-parameters related to the SSAE model,the QOAOA is derived by the including of the quasi-oppositional based learning(QOBL)with traditional AOA.Moreover,an objective function with the including of the integral of time multiplied by squared error(ITSE)is considered in this study.For validating the efficiency of the FOPID-QOAOA method,a sequence of simulations was performed under distinct aspects.A comparative study on cascaded buck and boost converters is carried out to authenticate the effectiveness and performance of the designed techniques.

Brain Tumor Diagnosis Using Sparrow Search Algorithm Based Deep Learning Model

Recently,Internet of Medical Things(IoMT)has gained considerable attention to provide improved healthcare services to patients.Since earlier diag-nosis of brain tumor(BT)using medical imaging becomes an essential task,auto-mated IoMT and cloud enabled BT diagnosis model can be devised using recent deep learning models.With this motivation,this paper introduces a novel IoMT and cloud enabled BT diagnosis model,named IoMTC-HDBT.The IoMTC-HDBT model comprises the data acquisition process by the use of IoMT devices which captures the magnetic resonance imaging(MRI)brain images and transmit them to the cloud server.Besides,adaptive windowfiltering(AWF)based image preprocessing is used to remove noise.In addition,the cloud server executes the disease diagnosis model which includes the sparrow search algorithm(SSA)with GoogleNet(SSA-GN)model.The IoMTC-HDBT model applies functional link neural network(FLNN),which has the ability to detect and classify the MRI brain images as normal or abnormal.Itfinds useful to generate the reports instantly for patients located in remote areas.The validation of the IoMTC-HDBT model takes place against BRATS2015 Challenge dataset and the experimental analysis is car-ried out interms of sensitivity,accuracy,and specificity.The experimentation out-come pointed out the betterment of the proposed model with the accuracy of 0.984.

Multiple Extreme Learning Machines Based Arrival Time Prediction for Public Bus Transport

Due to fast-growing urbanization,the traffic management system becomes a crucial problem owing to the rapid growth in the number of vehicles The research proposes an Intelligent public transportation system where informa-tion regarding all the buses connecting in a city will be gathered,processed and accurate bus arrival time prediction will be presented to the user.Various linear and time-varying parameters such as distance,waiting time at stops,red signal duration at a traffic signal,traffic density,turning density,rush hours,weather conditions,number of passengers on the bus,type of day,road type,average vehi-cle speed limit,current vehicle speed affecting traffic are used for the analysis.The proposed model exploits the feasibility and applicability of ELM in the travel time forecasting area.Multiple ELMs(MELM)for explicitly training dynamic,road and trajectory information are used in the proposed approach.A large-scale dataset(historical data)obtained from Kerala State Road Transport Corporation is used for training.Simulations are carried out by using MATLAB R2021a.The experiments revealed that the efficiency of MELM is independent of the time of day and day of the week.It can manage huge volumes of data with less human intervention at greater learning speeds.It is found MELM yields prediction with accuracy in the range of 96.7%to 99.08%.The MAE value is between 0.28 to 1.74 minutes with the proposed approach.The study revealed that there could be regularity in bus usage and daily bus rides are predictable with a better degree of accuracy.The research has proved that MELM is superior for arrival time pre-dictions in terms of accuracy and error,compared with other approaches.

Thermophoresis and Brownian motion effects on boundary layer flow of nanofluid in presence of thermal stratification due to solar energy

The problem of laminar fluid flow, which results from the stretching of a vertical surface with variable stream conditions in a nanofluid due to solar energy, is in- vestigated numerically. The model used for the nanofluid incorporates the effects of the Brownian motion and thermophoresis in the presence of thermal stratification. The sym- metry groups admitted by the corresponding boundary value problem are obtained by using a special form of Lie group transformations, namely, the scaling group of transfor- mations. An exact solution is obtained for the translation symmetrys, and the numerical solutions are obtained for the scaling symmetry. This solution depends on the Lewis number, the Brownian motion parameter, the thermal stratification parameter, and the thermophoretic parameter. The conclusion is drawn that the flow field, the temperature, and the nanoparticle volume fraction profiles are significantly influenced by these param- eters. Nanofluids have been shown to increase the thermal conductivity and convective heat transfer performance of base liquids. Nanoparticles in the base fluids also offer the potential in improving the radiative properties of the liquids, leading to an increase in the efficiency of direct absorption solar collectors.

Dust Deposition’s Effect on Solar Photovoltaic Module Performance:An Experimental Study in India’s Tropical Region

A solar PV panel works with maximum efficiency only when it is operated around its optimum operating point or maximum power point.Unfortunately,the performance of the solar cell is affected by several factors like sun direction,solar irradiance,dust accumulation,module temperature,as well as the load on the system.Dust deposition is one of the most prominent factors that influence the performance of solar panels.Because the solar panel is exposed to the atmosphere,dust will accumulate on its surface,reducing the quantity of sunlight reaching the solar cell and diminishing output.In the proposed work,a detailed investigation of the performance of solar PV modules is carried out under the tropical climatic condition of Chennai,India,where the presence of dust particles is very high.The data corresponding to four different dust samples of various densities at four solar irradiation levels of 220,525,702,and 905 W/m^(2)are collected,and performance analysis is carried out.Based on the analysis carried out,the maximum power loss is found to be 73.51%,66.29%,65.46%,and 61.42%,for coal,sand,brick powder,and chalk dust respectively.Hence,it can be said that coal dust contributes to the maximum power loss among all four dust samples.Due to heat dissipation produced by dust deposition,the performance of solar PV modules is degraded as the temperature rose.

Machine Learning-Based Pruning Technique for Low Power Approximate Computing

Approximate Computing is a low power achieving technique that offers an additional degree of freedom to design digital circuits.Pruning is one of the types of approximate circuit design technique which removes logic gates or wires in the circuit to reduce power consumption with minimal insertion of error.In this work,a novel machine learning(ML)-based pruning technique is introduced to design digital circuits.The machine-learning algorithm of the random forest deci-sion tree is used to prune nodes selectively based on their input pattern.In addi-tion,an error compensation value is added to the original output to reduce an error rate.Experimental results proved the efficiency of the proposed technique in terms of area,power and error rate.Compared to conventional pruning,proposed ML pruning achieves 32%and 26%of the area and delay reductions in 8*8 multi-plier implementation.Low power image processing algorithms are essential in various applications like image compression and enhancement algorithms.For real-time evaluation,proposed ML optimized pruning is applied in discrete cosine transform(DCT).It is a basic element of image and video processing applications.Experimental results on benchmark images show that proposed pruning achieves a very good peak signal-to-noise ratio(PSNR)value with a considerable amount of energy savings compared to other methods.

GSTP: Geographic Secured Two Phase Routing Using MD5 Algorithm

Secured Two Phase Geographic Greedy Forwarding (SecuTPGF) is a geographic greedy forwarding protocol for transmitting multimedia data stream in Wireless Multimedia Sensor Networks (WMSN) in a secure and reliable manner. Cryptographic and MAC authentication mechanisms are used to implement security for both node and message authentication. In this paper, a modified version of SecuTPGF, the GSTP routing provides security for both node and message authentication by using MD5 algorithm with a reduced computation power. In SecuTPGF, two different algorithms are used for node and message authentication, and GSTP routing uses “MD5Algorithm” for both node and message authentication. Using MD5 algorithm for node and message authentication, the average number of transmission paths increased and average number of hops used for transmission decreased when compared to the SecuTPGF. By conducting security analysis & evaluation experiments, the effectiveness of GSTP routing algorithm is proved.

Classification Using Two Layer Neural Network Back Propagation Algorithm

Worldwide breast cancer is the most common form of cancer death occurring in 12.6% of women. This paper presents a cost effective approach to classify the normal, malignant and benign tumor using two layer neural network back propagation algorithm. Back propagation algorithm is used to train the neural network. Parallelization techniques speed up the computation process and as a result two layer neural networks outperform the previous work in terms of accuracy. Breast cancer tumor database used for the testing purpose is from the CIA machine learning repository. The highest accuracy of 97.12% is achieved using the two layer neural network back propagation algorithm.

Detection of Link Fault Using Risk Modelling and Improving QOS Parameters in IP/WDM Networks

An operational backbone network is connected with many routers and other devices. Identifying faults in the network is very difficult, so a fault localization mechanism is necessary to identify fault and alleviate it and correct the faults in order to reduce the network performance degradation. A risk model needs to be devised based on the dynamic database by creating alternate path and the network is reconfigured by identifying dynamic paths. In this paper, an on-demand link state routing approach is used for handling failures in IP backbone networks and a localization algorithm is used to improve QOS parameters based on threshold value of gateway. It is proved that on-demand link state routing guarantees loop-free forwarding to reachable destinations regardless of the number of failures in the network, and in case of localization algorithm using modification process packet loss is avoided based on threshold value of gateway. Heuristic algorithm is also used for reconfiguration of dynamic path for effective fault localization. In this paper, in order to change the traffic condition, reconfiguration strategic is dynamically used. Dijikstra’s shortest path algorithm has been used to determine the shortest path between node pairs. Using risk modeling mechanism, a small set of candidate faults is identified. The concept of Fault Localization is used to minimize the fault occurring in the node and sends original path to node pairs. The localization algorithm based on MODIFICATION PROCESS, packet loss is avoided in the network by checking threshold value of gateway. If the threshold value is maximum, router directly forwards the packet to destination through gateway and if the threshold value is minimum, router compresses the packet and forwards the packet to destination with notification via gateway.

Electrical and non-linear optical studies on electrospun ZnO/BaO composite nanofibers

Nanocapacitors and nonvolatile ferroelectric random access memories require nanoscale thin film coatings with ferroelectric properties. One dimensional ferroelectric nanofibers are used in ferroelectric memory devices owing to the fact that decrease of the dimensionality of the memory device elements will reduce the addressing and appreciably increase the storage capacity, Novel ZnO/BaO nanocomposite fibers exhibiting ferroelectric properties have been prepared in the form of non-woven mesh by electrospinning the sol derived from the sol-gel route, Thin cylindrical nanofibers of average diameter 100 nm have been obtained and their morphology is confirmed by SEM and AFM images. In the electrospinning process, the effect of the working distance on the fiber morphology was studied and it showed that working distance between 11 and 15 cm can produce fibers without beads and the decrease in working distance in this range increases the fiber diameter. Powder XRD was used to identify the phases and EDX analysis confirmed the presence of ZnO/BaO. Dielectric and non-linear optical properties have also been studied. The dielectric studies showed that ZnO/BaO composite nanofibers undergo a phase transition from ferroelectric to paraelectric at 323 K.