Search

Providing Reinforcement Learning (RL) agents with human feedback can dramatically improve various aspects of learning. However, previous methods require human observer to give inputs explicitly (e.g., press buttons, voice interface), burdening the human in the loop of RL agent’s learning process. Further, providing explicit human advise (feedback) continuously is not always possible or too restrictive, e.g., autonomous driving, disabled rehabilitation, etc. In this work, we investigate capturing human’s intrinsic reactions as implicit (and natural) feedback through EEG in the form of error-related potentials (ErrP), providing a natural and direct way for humans to improve the RL agent learning. As such, the human intelligence can be integrated via implicit feedback with RL algorithms to accelerate the learning of RL agent. We develop three reasonably complex 2D discrete navigational games to experimentally evaluate the overall performance of the proposed work. And the motivation of using ErrPs as feedbacks is also verified by subjective experiments. Major contributions of our work are as follows, (i) we propose and experimentally validate the zero-shot learning of ErrPs, where the ErrPs can be learned for one game, and transferred to other unseen games, (ii) we propose a novel RL framework for integrating implicit human feedbacks via ErrPs with RL agent, improving the label efficiency and robustness to human mistakes, and (iii) compared to prior works, we scale the application of ErrPs to reasonably complex environments, and demonstrate the significance of our approach for accelerated learning through real user experiments.

The wireless sensor network (WSN) consist of battery-powered sensor nodes which are self-configured and are deployed for monitoring several physical or environmental conditions such as temperature, pressure, humidity, vibration, pollutants etc. The major constraint in most of the WSN applications is the replacement/recharging of the battery contained by the node once it gets exhausted. This limitation reduces the lifetime of WSN. The placement of energy harvesting device within the sensor node may be the best probable solution to recharge the exhausted battery. In this paper, the integration of low cost, light weight and foldable flexible solar cells with WSN has been focused. The aim of this paper is to fabricate the flexible solar cells and showing the potential use of them in WSN. Moreover, the use of flexible solar cell is the better selection for emerging wearable WSN. This paper also describes the various issues in the already developed energy harvesting models and suggested a self-powered model for energy management based on finite state machine (FSM). The proposed models completely avoid the overcharging and the frequent charging of the batteries. This optimal utilization of the battery maximizes the lifetime of WSN network. In the proposed model, the flexible p–i–n solar cells are used to convert solar energy into electrical energy that can charge the battery of the WSN node. Finally, it can be concluded that the node will continue to function actively till the battery lifetime i.e. approximately 25–30 years.

Wilson's disease (WD) is caused by copper accumulation in the brain and liver, and if not treated early, can lead to severe disability and death. WD has shown white matter hyperintensity (WMH) in the brain magnetic resonance scans (MRI) scans, but the diagnosis is challenging due to (i) subtle intensity changes and (ii) weak training MRI when using artificial intelligence (AI). Design and validate seven types of high-performing AI-based computer-aided design (CADx) systems consisting of 3D optimized classification, and characterization of WD against controls. We propose a "conventional deep convolution neural network" (cDCNN) and an "improved DCNN" (iDCNN) where rectified linear unit (ReLU) activation function was modified ensuring "differentiable at zero." Three-dimensional optimization was achieved by recording accuracy while changing the CNN layers and augmentation by several folds. WD was characterized using (i) CNN-based feature map strength and (ii) Bispectrum strengths of pixels having higher probabilities of WD. We further computed the (a) area under the curve (AUC), (b) diagnostic odds ratio (DOR), (c) reliability, and (d) stability and (e) benchmarking. Optimal results were achieved using 9 layers of CNN, with 4-fold augmentation. iDCNN yields superior performance compared to cDCNN with accuracy and AUC of 98.28 ± 1.55, 0.99 (p < 0.0001), and 97.19 ± 2.53%, 0.984 (p < 0.0001), respectively. DOR of iDCNN outperformed cDCNN fourfold. iDCNN also outperformed (a) transfer learning-based "Inception V3" paradigm by 11.92% and (b) four types of "conventional machine learning-based systems": k-NN, decision tree, support vector machine, and random forest by 55.13%, 28.36%, 15.35%, and 14.11%, respectively. The AI-based systems can potentially be useful in the early WD diagnosis. Graphical Abstract.

The problem of scheduling of tasks in distributed, heterogeneous, and multiprocessing computing environment like grid and cloud computing is considered as one of the most important issue from research perspective. As the performance of such kind of systems is highly depends upon the way, how tasks are allocated among the multiple processing units for their efficient execution. The underlying objective of any task scheduling mechanism is to minimize the overall makespan for the execution of given set of jobs/tasks and computing machines. Scheduling of tasks in cloud computing falls in the class of NP-hard optimization problem. As a result, many meta-heuristic algorithms have been applied and tested to solve this problem but still lot of scope is there for the better strategies. The characteristic of the good algorithm is that it must be adaptable to the dynamic environment. Through this paper, we are proposing task scheduling mechanism based on particle swarm optimization (PSO) in which opposition-based learning technique is used to avoid premature convergence and to accelerate the convergence of standard PSO and compared same with the well-established task scheduling strategies based on PSO, mPSO (modified PSO), genetic algorithm GA, max–min, minimum completion time and minimum execution time. The results obtained for the various class of experiments clearly establish that the proposed opposition-based learning inspired particle swarm optimization based scheduling strategy performs better in comparison to its peers which are taken into the consideration.

Wilson's disease (WD) is caused by the excessive accumulation of copper in the brain and liver, leading to death if not diagnosed early. WD shows its prevalence as white matter hyperintensity (WMH) in MRI scans. It is challenging and tedious to classify WD against controls when comparing visually, primarily due to subtle differences in WMH. This Letter presents a computer-aided design-based automated classification strategy that uses optimised transfer learning (TL) utilising two novel paradigms known as (i) MobileNet and (ii) the Visual Geometric Group-19 (VGG-19). Further, the authors benchmark TL systems against a machine learning (ML) paradigm. Using four-fold augmentation, VGG-19 is superior to MobileNet demonstrating accuracy and area under the curve (AUC) pairs as 95.46 ± 7.70%, 0.932 (p < 0.0001) and 86.87 ± 2.23%, 0.871 (p < 0.0001), respectively. Further, MobileNet and VGG-19 showed an improvement of 3.4 and 13.5%, respectively, when benchmarked against the ML-based soft classifier – Random Forest.

Background & Objective: Cloud computing emerges out as a new way of computing which enables the users to fulfill their computation need using the underlying computing resources like software, memory, computing nodes or machines without owning them purely on the basis of pay-per-use that too round the clock and from anywhere. People defined this as the extension of the existing technologies like parallel computing, distributed computing or grid computing. Lots of research have been conducted in the field of cloud computing but the task scheduling is considered to be the most fundamental problem which is still in infancy and requires a lot of attention and a proper mechanism for the optimal utilization of the underlying computing resources. Task scheduling in cloud computing environment lies into the category of NP-hard problem and many heuristics and Meta heuristics strategies have been applied to solve the problem. Methods: In this work, Fuzzy Enabled Genetic Algorithm (FEGA) is proposed to solve the problem of task scheduling in cloud computing environment as classical roulette wheel selection method has certain limitations to solve complex optimization problem. Results & Discussion: In this work, an efficient fuzzy enabled genetic algorithm based task scheduling mechanism has been designed, implemented and investigated. The efficiency of the proposed FEGA algorithm is tested using various randomly generated data sets in different situations and compared with the other meta-heuristics. Conclusion: The authors suggest that the proposed Fuzzy Enabled Genetic Algorithm (FEGA) to solve the task scheduling problem helps in minimizing the total execution time or makespan and on comparing with other Meta-heuristic like genetic algorithm and greedy based strategy found that FEGA outperforms the both in different set of experiments.

The numerical simulation of lead-free all-perovskite tandem solar cell using solar capacitance one dimensional tool (SCAPS 1D) has been performed. The combination of methyl ammonium germanium halide (CH3NH3GeI3) as wide band-gap (1.9 eV) perovskite layer and FA0.75MA0.25Sn0.25 Ge0.5I3 (FAMASnGeI3) as a low band-gap (1.4 eV) perovskite layer is considered for the first time which shows the optimal device efficiency of 26.72%. The present work studies the effect of the different electron transport layer, perovskite absorber thickness and its defect density on photovoltaic performance. Further, the effect of front electrode work function on the device photovoltaic performance is also studied. It has been observed that the device open circuit voltage (VoC) is significantly affected by the built-in voltage (Vbi) across the perovskite layer. In addition, the effect of thickness of the perovskite absorber layers is also investigated. Moreover, the effect of defect density of the absorber layer is also explored, and it has been found that for better device performance the defect density needs to be as low as possible. Furthermore, it has been observed that the work function of the front contact plays a significant role and for the proposed device its value should not be more than 4.4 eV. The proposed all–perovskite tandem solar cell shows the remarkable photovoltaic parameters i.e., open circuit voltage (VOC) = 1.07 Volts, short circuit current density (JSC) = 28.36 mA/cm2, fill factor (FF) = 84.39% and efficiency (η) = 26.72%.

Tomato is the most popular crop in the world and in every kitchen, it is found in different forms irrespective of the cuisine. After potato and sweet potato, it is the crop which is cultivated worldwide. India ranked 2 in the production of tomato. However, the quality and quantity of tomato crop goes down due to the various kinds of diseases. So, to detect the disease a deep learning-based approach is discussed in the article. For the disease detection and classification, a Convolution Neural Network based approach is applied. In this model, there are 3 convolution and 3 max pooling layers followed by 2 fully connected layer. The experimental results shows the efficacy of the proposed model over pre-trained model i.e. VGG16, InceptionV3 and MobileNet. The classification accuracy varies from 76% to 100% with respect to classes and average accuracy of the proposed model is 91.2% for the 9 disease and 1 healthy class.

In this paper, enhancement of volume depletion is studied on P-type double gate junctionless Field Effect Transistor (P-DGJLFET) by Gate work function and Gate dielectric engineering. The formation...

Cloud computing is an emerging technology which involves the allocation and de-allocation of the computing resources using the internet. Task scheduling (TS) is one of the fundamental issues in cloud computing and effort has been made to solve this problem. An efficient task scheduling mechanism is always needed for the allocation to the available processing machines in such a manner that no machine is over or under-utilized. Scheduling tasks belongs to the category of NP-hard problem. Through this article, the authors are proposing a particle swarm optimization (PSO) based task scheduling mechanism for the efficient scheduling of tasks among the virtual machines (VMs). The proposed algorithm is compared using the CloudSim simulator with the existing greedy and genetic algorithm-based task scheduling mechanism. The simulation results clearly show that the PSO-based task scheduling mechanism clearly outperforms the others as it results in almost 30% reduction in makespan and increases the resource utilization by 20%.

Background COVID-19 pandemic has currently no vaccines. Thus, the only feasible solution for prevention relies on the detection of COVID-19-positive cases through quick and accurate testing. Since artificial intelligence (AI) offers the powerful mechanism to automatically extract the tissue features and characterise the disease, we therefore hypothesise that AI-based strategies can provide quick detection and classification, especially for radiological computed tomography (CT) lung scans. Methodology Six models, two traditional machine learning (ML)-based (k-NN and RF), two transfer learning (TL)-based (VGG19 and InceptionV3), and the last two were our custom-designed deep learning (DL) models (CNN and iCNN), were developed for classification between COVID pneumonia (CoP) and non-COVID pneumonia (NCoP). K10 cross-validation (90% training: 10% testing) protocol on an Italian cohort of 100 CoP and 30 NCoP patients was used for performance evaluation and bispectrum analysis for CT lung characterisation. Results Using K10 protocol, our results showed the accuracy in the order of DL > TL > ML, ranging the six accuracies for k-NN, RF, VGG19, IV3, CNN, iCNN as 74.58 ± 2.44%, 96.84 ± 2.6, 94.84 ± 2.85%, 99.53 ± 0.75%, 99.53 ± 1.05%, and 99.69 ± 0.66%, respectively. The corresponding AUCs were 0.74, 0.94, 0.96, 0.99, 0.99, and 0.99 (p-values

In proposed work, a compressed version of UNet has been developed using Differential Evolution for segmenting the diseased regions in leaf images. The compressed model has been evaluated on potato late blight leaf images from PlantVillage dataset. The compressed model needs only 6.8% of space needed by original UNet architecture, and the inference time for disease classification is twice as fast without loss in performance metric of mean Intersection over Union (IoU).

SegNet is a Convolution Neural Network (CNN) architecture consisting of encoder and decoder for pixel-wise classification of input images. It was found to give better results than state of the art pixel-wise segmentation of images. In proposed work, a compressed version of SegNet has been developed using Differential Evolution for segmenting the diseased regions in leaf images. The compressed model has been evaluated on publicly available street scene images and potato late blight leaf images from PlantVillage dataset. Using the proposed method a compression of 25x times is achieved on original SegNet and inference time is reduced by 1.675x times without loss in mean IOU accuracy.

Line-of-sight (LOS) is a critical requirement for mmWave wireless communications. In this work, we explore the use of access point (AP) infrastructure mobility to optimize indoor mmWave WiFi network performance based on the discovery of LOS connectivity to stations (STAs).We consider a ceiling-mounted mobile (CMM) AP as the infrastructure mobility framework. Within this framework, we present a LOS prediction algorithm based on machine learning (ML) that addresses the LOS discovery problem. The algorithm relies on the available network state information (e.g., LOS connectivity between STAs and the AP) to predict the unknown LOS connectivity status between the reachable AP locations and target STAs. We show that the proposed algorithm can predict LOS connectivity between the AP and target STAs with an accuracy up to 91%. Based on the LOS prediction algorithm, we then propose a systematic solution WiMove, which can decide if and where the AP should move to for optimizing network performance. Using both ns-3 based simulation and experimental prototype implementation, we show that the throughput and fairness performance of WiMove is up to 119% and 15% better compared with single static AP and brute force search.

Task scheduling is one of the most difficult problems which is associated with cloud computing. Due to its nature, as it belongs to nondeterministic polynomial time (NP)-hard class of problem. Various heuristic as well as meta-heuristic approaches have been used to find the optimal solution. Task scheduling basically deals with the allocation of the task to the most efficient machine for optimal utilization of the computing resources and results in better makespan. As per literature, various meta-heuristic algorithms like genetic algorithm (GA), particle swarm optimization (PSO), ant colony optimization (ACO) and their other hybrid techniques have been applied. Through this paper, we are presenting a novel meta-heuristic technique — genetic algorithm enabled particle swarm optimization (PSOGA), a hybrid version of PSO and GA algorithm. PSOGA uses the diversification property of PSO and intensification property of the GA. The proposed algorithm shows its supremacy over other techniques which are taken into consideration by presenting less makespan time in majority of the cases which leads up to 22.2% improvement in performance of the system and also establishes that proposed PSOGA algorithm converges faster than the others.

This paper deals with the engineering of the hetero-interface between intrinsic amorphous silicon (i-a-Si: H) layer and the n-type crystalline silicon (c-Si) wafer during the fabrication of the Silicon Heterojunction (SHJ) solar cell by the Hot Wire Chemical Vapor Deposition technique. It is known that this interface and the associated surface passivation of the c-Si is key to obtaining high efficiency heterojunction solar cells. The monitoring of this interface was carried out using high-resolution transmission electron microscopy (HRTEM). The HRTEM data of the c-Si/a-Si:H interface reveals a drastic dependence on the filament temperature (T-f) used during the deposition of the i-a-Si:H layer. Detailed analysis of the solar cell characteristics indicates that the cells where one has an abrupt crystalline/amorphous interface shows higher conversion efficiency compared to those where we have a rough and a defective interface or where there are indications of local epitaxy in the a-Si:H layer. The second parameter which was engineered is the bulk defect density of the intrinsic a-Si:H layer. Though the thickness of i-a-Si:H layer in case of SHJ solar cells is only around 5 nm and serves the purpose of passivating the dangling bonds on the c-Si wafer, the bulk defect density of this layer cannot be ignored. We have achieved a-Si:H films with acceptable bulk defect density without dilution of the silane gas with hydrogen. The bulk defect density of the i-a-Si:H layer has been determined by the constant photocurrent method (CPM) and is correlated to the performance of SHJ solar cells. A direct consequence of these control parameters was observed in the improvement of the external quantum efficiency near 600 nm wavelength region. (C) 2017 Elsevier Ltd. All rights reserved.

In the proposed work, a convolution neural network (CNN) based model has been used to identify the cancer disease in human lung(s). Moreover, this approach identifies the single or multi-module in lungs by analyzing the Computer Tomography (CT) scan. For the purpose of the experiment, publicly available dataset named as Early Lung Cancer Action Program (ELCAP) has been used. Moreover, the performance of proposed CNN model has been compared with traditional machine learning approaches i.e. support vector machine, k-NN, Decision Tree, Random Forest, etc under various parameters i.e. accuracy, precision, recall, Cohen Kappa. The performance of proposed model is also compared with famous CNN models i.e. VGG16, Inception V3 in terms of accuracy, storage space and inference time. The experimental results show the efficacy of proposed algorithms over traditional machine learning and pre-trained models by achieving the accuracy of 99.5%

Presently, the prediction of share is a challenging issue for the research community as share market is a chaotic place. The reason behind it, there are several factors such as government policies, international market, weather, performance of company. In this article, a model has been developed using long short term memory (LSTM) to predict the share price of DLF group. Moreover, for the experimental purpose the data of DLF group has been taken from yahoo financial services in the time duration of 2008 to 2018 and the recurrent neural network (RNN) model has been trained using data ranging from 2008 to 2017. This RNN based model has been tested on the data of year 2018. For the performance comparison purpose, other linear regression algorithms i.e. k-nn regression, lasso regression, XGboost etc has been executed and the proposed algorithm outperforms with 2.6% root mean square error.

The agricultural production is affected by the climate changes i.e. humidity, rain, extremes of temperature etc. Additionally, abiotic stresses are causative element to the etiology of disease as well as pest on crops. The production of the crops can be improved by diagnosis as well as detecting the accurate disease on time or in early stage. Moreover, it is very difficult for accurately detecting and treatment based on the technique which used in disease and insect pests diagnosis. Few researchers have made efforts on predicting disease as well as pest crops using machine learning algorithms. Therefore, this paper presents disease identification in corn crops by analyzing the leaves in the very early stage. We have used PlantVillage dataset for experiments and analysis. The validity of the results has been cheeked on various performance metrics such as precision, accuracy, recall, storage space, running time of the model and AUC-RoC. The obtained results shows the proposed technique outperform in comparison with the traditional machine learning algorithms. Developed model is able to achieve the accuracy of 94%.

In this paper, we present a CNN model to identify the disease in grapes plant in early stage by analyzing the leaf images. The model is illustrated on publicly available PlantVillage dataset and the performance compared with traditional machine learning models and pre-trained convolution neural network models. The performance of algorithms has been compared on various evaluation metrics i.e. accuracy, precision recall, running time of model, storage space and AUC-RoC. From the experimental results, it has been observed that the performance of proposed model is better than traditional machine learning algorithms as well as pre-trained models by achieving an accuracy of 99%.

Eye-blinks are known to substantially contaminate EEG signals, and thereby severely impact the decoding of EEG signals in various medical and scientific applications. In this work, we consider the problem of eye-blink detection that can then be employed to reliably remove eye-blinks from EEG signals. We propose a fully automated and unsupervised eyeblink detection algorithm, Blink that self-learns user-specific brainwave profiles for eye-blinks. Hence, Blink does away with any user training or manual inspection requirements. Blink functions on a single channel EEG, and is capable of estimating the start and end timestamps of eye-blinks in a precise manner. We collect four different eye-blink datasets and annotate 2300+ eye-blinks to evaluate the robustness performance of Blink across headsets (OpenBCI and Muse), eye-blink types (voluntary and involuntary), and various user activities (watching a video, reading an article, and attending to an external stimulation). The Blink algorithm performs consistently with an accuracy of over 98% for all the tasks with an average precision of 0.934. The source code and annotated datasets are released publicly for reproducibility and further research. To the best of our knowledge, this is the first ever annotated eye-blink EEG dataset released in the public domain.

Automated primate face recognition has enormous potential in effective conservation of species facing endangerment or extinction. The task is characterized by lack of training data, low inter-class variations, and large intra-class differences. Owing to the challenging nature of the problem, limited research has been performed to automate the process of primate face recognition. In this research, we propose a novel Triplet Transform Learning (TTL) model for learning discriminative representations of primate faces. The proposed model reduces the intra-class variations and increases the inter-class variations to obtain robust sparse representations for the primate faces. It is utilized to present a novel framework for primate face recognition, which is evaluated on the primate dataset, comprising of 80 identities including monkeys, gorillas, and chimpanzees. Experimental results demonstrate the efficacy of the proposed approach, where it outperforms the existing approaches and attains state-of-the-art performance on the primates database.

Videos are one of the most engaging and interesting mediums of effective information delivery and constitute the majority of the content generated online today. As human attention span shrinks, it is imperative to shorten videos while maintaining most of its information. The premier challenge is that summaries more intuitive to a human are difficult for machines to generalize. We present a simple approach to video summarization using Kernel Temporal Segmentation (KTS) for shot segmentation and a global attention based modified memory network module with LSTM for shot score learning. The modified memory network termed as Global Attention Memory Module (GAMM) increases the learning capability of the model and with the addition of LSTM, it is further able to learn better contextual features. Experiments on the benchmark datasets TVSum and SumMe show that our method outperforms the current state of the art by about 15%.

Fruits are common items bought by every household. They are delicious to eat and rich in nourishment. However they may also adversely affect health if the fruits are from a diseased tree/plant. Moreover, Farmers may also loose lot of amount of profit if their plants get affected by some disease. In this article, the main objective/goal is to develop a convolution neural network based approach to identify the disease in apple fruit. The data for experiment has been taken from PlantVillage. In the proposed work, a convolution neural network model has been developed to identify the disease in apple and it consists of three convolution layer, three max pooling layer followed by two densely connected layers. This model was formed after testing with varying number of convolution layers from 2 to 6 and found that 3 layer was giving best accuracy. For the result comparison purpose, the traditional machine learning algorithms are also executed on the same dataset. Along with traditional machine learning approaches, the famous pre-trained CNN models i.e. VGG16 and InceptionV3 are also executed. The experiments results shows the efficacy of proposed algorithm over pre-trained models and traditional machine learning approach in terms of accuracy, computational time, specificity, F1 score and AUC-ROC curve. The proposed model achieves the state of the art accuracy of 99%. Moreover, the proposed model requires only 20% of the space as compared to pre-trained model with inference time less than 1 second as pre-trained models require minimum 30 second.

Deforestation and loss of habitat have resulted in rapid decline of certain species of primates in forests. On the other hand, uncontrolled growth of a few species of primates in urban areas has led to safety issues and nuisance for the local residents. Hence, identifying individual primates has become the need of the hour - not only for conservation and effective mitigation in the wild but also in zoological parks and wildlife sanctuaries. Primates and human faces share a lot of common features like position and shape of eyes, nose and mouth. It is worth exploring whether the knowledge of human faces and recent methods learned from human face detection and recognition can be extended to primate faces. However, similar challenges relating to bias in human faces will also occur in primates. The quality and orientation of primate images along with different species of primates - ranging from monkeys to gorillas and chimpanzees will contribute to bias in effective detection and recognition. Experimental results on a primate dataset of over 80 identities show the effect of bias in this research problem.

Cucumber is one of the important crop and farmers of most of the counties are cultivating the cucumber crop. Generally, this crop is infected with Angular Spot, Anthracnose, etc. In past research community has developed various learning models to identify the disease in cucumber crop in early-stage and reported maximum accuracy of 85.7%. In proposed work, a convolution neural network based approach has been discussed and disease identification is improved by 8.05% by achieving the accuracy of 93.75%. The proposed model has been trained on a different combination of hyperparameters and activation function. However, the best accuracy has been achieved by introducing a modified ReLU activation function. A segmentation algorithm has also been proposed to estimate the severity of the disease. To establish the efficacy of the proposed model, its performance has been compared with other CNN models as well as traditional machine learning methods.

This article presents the numerical simulation of lead–free (Pb-free) double–perovskite (Cs2AgBi0.75Sb0.25Br6, i.e., mixed antimony bismuth halide double-perovskite) solar cell (DPSC) using solar cell capacitance one dimensional simulator (SCAPS-1D). In this contribution, the hole transport layers (HTLs) and electron transport layers (ETLs) are optimized. The optimization is performed by correlating the open circuit voltage (VOC) with the built-in potential (Vbi). It is disclosed from the simulation results, that higher Vbi resulted higher VOC. Furthermore, it is also found that, for the proper selection of HTL, the EV_HTL (Valence band maximum of HTL) and φBC (Work function of back contact) should not be much deeper than the EV_PVK (Valence band maximum of double–perovskite layer) to avoid Vbi loss. Correspondingly, for proper selection of ETL, EC_ETL (Conduction band minimum of electron transport layer) and ΦFC (Work function of front contact), should not be much higher than EC_PVK to prevent Vbi loss. The optimized HTL and ETL are found to be Cu2O (copper(I) oxide) and ZnOS (Zinc oxysulfide) respectively. Moreover, the device photovoltaic performance is further improved by optimizing the double-perovskite layer thickness which is found to be 400 nm. Under optimized condition, the device photovoltaic power conversion efficiency (η) improves to 18.18%. The optimized device photovoltaic performance are found to be, open –circuit voltage (Voc) = 1.39 V, short-circuit current density (JSC) = 16.04 mA/cm2, and fill factor (FF) = 78.34% which indicates that double–perovskite absorber layer (Cs2AgBi0.75Sb0.25Br6) is a suitable candidate for the development of highly efficient Pb-free DPSC.

The numerical simulation of lead-free perovskite cesium tin germanium halide (CsSnGeI3) and methyl ammonium germanium halide (CH3NH3GeI3) solar cells have been performed. It has been elucidated from the numerical simulation that as compare to the CH3NH3GeI3 based structure, CsSnGeI3 based structure depicts better photovoltaic performance under constant illumination condition. To enhance the device photovoltaic performance, the effects of different hole transport layer (HTL), defect density of the absorber layer, metal work function and the effect of temperature has been studied. In the present investigation, the effect of hole transport layers are analyzed by correlating the built-in voltage (Vbi) with the open circuit voltage (VOC). It has been obtained from the simulation results, that Vbi has a significant impact on VOC as the higher Vbi corresponds to the highest VOC. Moreover, the effect of defect density and metal work function is also studied. It is obtained from the simulation results, that the optimum defect density is found to be 1 × 1014 cm−3 and metal work function should be greater than or equal to 5 eV for both lead-free CH3NH3GeI3 and CsSnGeI3 based perovskite layer. Furthermore, to suggest possible lead-free alternatives for all-perovskite two-terminal multi-junction solar cell, an attempt is made to match the current of bottom perovskite layer i.e., CsSnGeI3 based perovskite layer to the top perovskite layer i.e., CH3NH3GeI3 based perovskite layer. To attain the current matching for bottom perovskite layer (CsSnGeI3 based perovskite layer) to top perovskite layer (CH3NH3GeI3 based perovskite layer), the bottom perovskite sub cell is fed with the filtered spectrum which is obtained by transmission from top electrode. After, the realization of current matching the short circuit current density of CsSnGeI3 based perovskite layer drops from 25.75 mA/cm2 to 15.32 mA/cm2 which is similar to CH3NH3GeI3 based perovskite absorber layer. In addition, it has been obtained from literature, that both the perovskite layers can be fabricated using solution–processing low-temperature technology. Hence, the numerical study suggests the possible alternatives for wide and narrow bandgap perovskite layers for the application to obtain highly efficient lead-free all-perovskite multi-junction solar cells.

The effect of substrate temperature on the structural and electrical properties of multiferroic Bi0.90La0.10Fe0.95Mn0.05O3 (BLFMO) thin films deposited on Pt (111)/Ti/SiO2/Si substrate using pulsed laser deposition (PLD) has been investigated. Films with substrate temperature ranging from 450 degrees C to 650 degrees C have been deposited. The grain size and roughness are found to increase with substrate temperature. The film deposited at 575 degrees C exhibitsmaximumremnant polarization around 39 mu C/cm(2) and a coercive field of 400 kV/cm. The cyclic fatigue study of the sample shows only 4% loss after 10(8) cycles. Complex impedance study of the BLFMO thin films demonstrates electrical homogeneity of the sample. AC conductivity data has been fitted using Jonscher single power law, and value of n found to be

Surface of n-type emitter solar cell is subjected to Oleylamine coating using standard chemical methods. The solar cell shows an increment by 16.8% in the efficiency slowly degrading in time. Investigations are carried out to isolate the passivating effect as well as degrading reactions. The degradation manifests in the form of increased reflectivity of the surface. It is expected that appropriate choice of the coatings can lead to efficient passivation of solar cell with low heat budget over extended time periods suitable for commercial applications. Our study provides a basis for processing the Si-solar cells for room temperature passivation. (C) 2016 Elsevier Ltd. All rights reserved.

Character recognition is an important problem in Pattern Classification. It is difficult to recognize the character in images or pdf or scanned text or handwritten characters scan. There are various ways to recognize text which comprises of Neural Network, Genetic Algorithm, Soft computing and fuzzy logic. In this paper, we compare the various ways already explored by researchers. We have tried to compare the Artificial Neural Network and Genetic Algorithm for the character recognition. The character image is initially segmented into pixel array and then normalized and skew removed with feature extraction to form an input to the neural network or the genetic algorithm. We work with two set of data, training data and test data and the aim is to recognize character correctly in test data. General Terms Pattern Recognition, Review, Image Processing, Artificial Neural Network, Genetic Algorithm et. al.

Tomato is an important vegetable crop cultivated worldwide coming next only to potato. However, the crop can be damaged due to various diseases. It is important for the farmer to know the type of disease for timely treatment of the crop. It has been observed that leaves are clear indicator of specific diseases. A number of Machine Learning (ML) algorithms and Convolution Neural Network (CNN) models have been proposed in literature for identification of tomato crop diseases. CNN models are based on Deep Learning Neural Networks and differ inherently from traditional Machine Learning algorithms like k-NN, Decision-Trees etc. While pretrained CNN models perform fairly well, they tend to be computationally heavy due to large number of parameters involved. In this paper a simplified CNN model is proposed comprising of 8 hidden layers. Using the publicly available dataset PlantVillage, proposed light weight model performs better than the traditional machine learning approaches as well as pretrained models and achieves an accuracy of 98.4%. PlantVillage dataset comprises of 39 classes of different crops like apple, potato, corn, grapes etc. of which 10 classes are of tomato diseases. While traditional ML methods gives best accuracy of 94.9% with k-NN, best accuracy of 93.5% is obtained with VGG16 in pretrained models. To increase performance of proposed CNN, image pre-processing has been used by changing image brightness by a random value of a random width of image after image augmentation. The proposed model also performs extremely well on dataset other than PlantVillage with accuracy of 98.7%.

A rectangular core is inserted in double gate junctionless transistor (DGJLT) which separates the top shell and bottom shell in the device called as rectangular core–shell double gate junctionless transistor (RCS-DGJLT). The core doping and core thickness are optimized to improve the performance of RCS-DGJLT. The core thickness is optimized for different shell thicknesses in the device which has not been explored yet in the literature. An interesting observation is found that the core thickness should be equal to the shell thickness for smaller shell thicknesses, whereas the thicker core is required for larger shell thicknesses. RCS-DGJLT has shown remarkable improvement than DGJLT. Further, the effect of gate misalignment on either side (source/drain) of the channel in RCS-DGJLT and DGJLT is studied and compared on the basis of device performance. The study of gate misalignment becomes essential due to the complications on achieving perfectly aligned gates during the fabrication of double gate device. An RCS-DGJLT is found to have better tolerance to gate misalignment than DGJLT. The best parametric values like OFF current is ~10–16A, ON current is ~10–5A, subthreshold slope is nearly 66.2 mV/decade, ON/OFF current ratio is ~1010, drain induced barrier lowering is ~42.1 mV V−1 and threshold voltage ~0.56 V at perfectly aligned gate condition is obtained on keeping core and shell thickness at 3 nm each. As per the analysis, the gate misalignment up to 20% on either side of the channel does not impact the performance parameters much and hence reduces the stress of meeting the requirement of perfect gate alignment.

Data warehouse is the most important asset of an organization as it contains highly valuable and sensitive information that is useful in decision-making process. The data warehouse provides easy access to organizational data as it contains data from different sources. Thus, it is essential to structure security measures for the protection of data that resides in data warehouse against malicious attackers to ensure proper security and confidentiality. Security should be considered vital from initial stages of designing data warehouse and hence should be deployed. Though a lot of work is being done towards the improvement and development of data warehouse till now but very less attention is given on the implementation of the security approaches in data warehouse. This paper focuses on improving the level of security by combining One-Time Pad (OTP) encryption technique with Advanced Encryption Standard (AES) to encrypt the data before loading it into data warehouse. Finally, the proposal of the model is to incorporate OTP encryption technique in the architecture of data warehouse for enhancing its security.

Cloud computing is one of the most acceptable emerging technologies, which involves the allocation and de-allocation of the computing resources using the Internet as the core technology to compute the tasks or jobs submitted by the users. Task scheduling is one of the fundamental issues in cloud computing and lots of efforts have been made to solve this problem. For the success of any cloud-based computing model, efficient task scheduling mechanism is always needed which, in turn, is responsible for the allocation of tasks to the available processing machines in such a manner that no machine is over- or under-utilized while executing them. Scheduling of tasks belongs to the category of NP-Hard problem. Through this paper, we are proposing the particle swarm optimization (PSO)-based task scheduling mechanism for the efficient distribution of the task among the virtual machines (VMs) in order to keep the overall response time minimum. The proposed algorithm is compared using the CloudSim simulator with the existing greedy and genetic algorithm-based task scheduling mechanism and results clearly shows that the PSO-based task scheduling mechanism clearly outperforms the others techniques which are taken into consideration.

Wireless sensor network (WSN) is widely used in a variety of applications including habitat monitoring, military surveillance, environmental monitoring, scientific applications, etc. The major limitation of WSN is that sometimes it is not feasible to replace or recharge the battery once it gets fully exhausted and thus, it limits the lifetime of WSN. One of the possible solutions to overcome this limitation is to incorporate any energy harvesting device, which can use the alternative energy sources to charge the battery. However, the processing temperature and the performance of energy harvesting devices limit their applications. In this paper, low temperature and high performance single-sided silicon heterojunction (SHJ) solar cells are fabricated with 13% efficiency using hot-wire chemical vapor deposition (HWCVD) method. This paper also describes an energy management model that successfully addresses the various issues in the existing energy harvesting models. In order to implement the proposed model, the results show that the high efficiency SHJ solar cells are best suitable candidate as an energy harvesting device that can be incorporated inside the node. The subsequent analysis shows that the consumed power per day by the node can be successfully recovered from the SHJ solar cells, if the sunlight is available only for 25 min in a day with 100 mW/cm2 intensity. This clearly indicates that the node's battery will remain fully charged if the above said condition is satisfied, which seems to be very feasible. Finally, one can conclude that the node functioning will remain active till the battery lifetime i.e., approximately 30 years for Li-ion battery.

In the current scenario data security has become an important issue with the growth of digital media. Many users and the applications are accessing the data both from inside and outside the database. Hence, the database as well as data within these databases has become the key target for most of the attackers. Many cryptographic schemes have been designed to solve this problem. Encryption plays an important role in providing the data confidentiality to data stored within the databases. But, the problem in adopting the standard encryption methods is that they may cause a damage to the existing schema as well as to the underlying applications or database as the output length is different from the input length and it also changes the format of data. This paper proposes a Format Preserving Encryption method by accumulating with Advance encryption standard(AES), eXclusive OR operation and a translation method for 16 digit numeric data. Format preserving encryption technique is used to minimizes the databases changes by preserving the format as well as the length of the input data.

The Parity Generators and checkers are crucial block in Central Processing Unit especially in the process of Error Detection and correction during data transmission. In this work, the three-bit Quantum Cellular Automata (QCA) even parity generator have been proposed using Layered T AND and OR Gate. The proposed comparator layout needs 9.02% less effective area compared to the best reported design so far. Moreover, the other QCA design metrics such as Effective Area, O-Cost, Costa, Irreversible Power Dissipation, and Complexity are analyzed for the proposed Parity Generator circuit. The functionality of the proposed circuit is verified by computer aided design tool QCADesigner.

Recently, Cloud computing emerges out as a latest technology which enables an organization to use the computing resources like hardware, applications, and software, etc., to perform the computation over the internet. Cloud computing gain so much attention because of advance technology, availability, and cost reduction. Task scheduling in cloud computing emerges out as new area of research which attracts the attention of lots researchers. An effective task scheduling is always required for optimum or efficient utilization of the computing resources to avoid the situation of over or under-utilization of such resources. Through this paper, we are going to proposed the cuckoo search-based task scheduling approach which helps in distributing the tasks efficiently among the available virtual machines (VM’s) and also keeps the overall response time (QoS) minimum. This algorithm assigns the tasks among the virtual machines on the basis of their processing power, i.e., million instructions per seconds (MIPS) and length of the tasks. A comparison of cuckoo search algorithm is done with the first—in first—out (FIFO) and greedy-based scheduling algorithm which is performed using the CloudSim simulator, the results clearly shows that cuckoo search outperforms the other algorithms.

Objective The current work was attempted to formulate and evaluate the topical sustained release delivery systems called cubosomes containing Tretinoin for the acne treatment. The recent patents on various formulations of tretinoin (EP0408370A2) helped in selecting a new formulation and evaluation. Methods Tretinoin loaded cubosomes were prepared by bottom-up technique, using varying the concentration of lipid and surfactant and keeping the drug concentration constant, a total of nine formulations of tretinoin was developed. These preparations were evaluated for surface charge, particle size, particle morphology, encapsulation efficiency, in-vivo and in-vitro release studies of gel enriched with cubosome dispersion. Finally, the stability studies of cubosomal gel were performed on optimized formulations. Results Significant result were obtained with tretinoin formulation as the drug is lipophilic, so it gives more depot effect on the epidermis and good retention property. The data obtained from the formulations showed that formulation TCF-5 was the optimized formulation which exhibited better drug release and entrapment efficiency. Conclusion It can be concluded that cubosomes offer benefits of quick onset as well as the maximal release of drug with fewer side effects. Thus, cubosomes represent a capable transporter having the property to sustain the release of drug, potential to localize the drug in the skin with a possible clinical application for acne vulgaris treatment due to cubosome depot effect on the epidermis.

Brain-computer interfaces (BCIs) allow users to communicate to a nearby computing device (computer, smartphone, etc.) using thoughts or other covert actions that result in a detectable change in brain-waves. Consider a BCI command to be a word consisting of a sequence of characters. Each character is a thought or action that can be reliably detected through brain waves. For this work, we specifically consider eye-blinks as the user action of interest. Eye-blinks are an interesting modality for BCI commands because of their easy detectability and naturalness (and hence covertness). It turns out that there is an interesting trade-off between the complexity of characters and the length of words. In this work, we perform a user-study to answer a simple, but important, question pertaining to eye-blinks based BCI command design: do users prefer shorter characters (and hence longer words) or shorter words (and hence longer characters) when performing commands?. We present a simple eye-blink language consisting of words and characters and use real user-experiments to study the aforementioned trade-off.

Automatic modulation classification (AMC) with multiple sensors is a challenging problem when the channel conditions are unknown at the receiver. In this paper, using the Markov chain Monte Carlo (MCMC) approach, we develop a novel algorithm for AMC when the amplitude and phase of the channel gains are unknown. Using sampling techniques, we marginalize over the unknown channel parameters that follow a certain probability distribution. This improves the estimate of the a posteriori distribution of the modulation formats, thereby improving the overall classification performance. Further, to overcome the problem of local extrema traps encountered in sampling algorithms, we introduce the idea of adding artificial noise beyond a certain threshold of signal-to-noise (SNR). This improves the performance of the sampling based AMC algorithm in the high SNR regime. Simulation results and comparisons are provided to show the efficiency of the proposed algorithm over the most related works in the literature.

Multiferroic materials have attracted the attention of scientific community all over the world due to smart functionality and attractive applications such as in magnetoelectric memory and magnetoelectric sensors. In this paper, we report on the ferroelectric, magnetic and magnetoelectric properties of microwave sintered BiFeO3 (BFO) and Bi0.90La0.10Fe0.95Mn0.05O3 (BLFMO) ceramics. The ferromagnetic and paramagnetic components have been deconvoluted using standard magnetization equation. The ferromagnetic saturation and paramagnetic susceptibility for BFO are found to be 0.046 emu/g and 1.17 x 10(-5) emu/g.Oe respectively. The magnetization in BLFMO increases enormously as compared to BFO due to suppression of cycloid spin. The magnetoelectric voltage coefficient for BFO is 3.37 mV/cm.Oe, while for BLFMO it is 10.11 mV/cm.Oe. As compared to BFO, BLFMO exhibits improved ferroelectric, magnetic and magnetoelectric properties. (C) 2017 Elsevier B.V. All rights reserved.

In the current scenario, Cloud computing carved itself as an emerging technology which enables the organization to utilize hardware, software and applications without any upfront cost over the internet. The challenge before the cloud service provider is, how efficiently and effectively the underlying computing resources like virtual machines, network, storage units, and bandwidth etc. should be managed so that no computing device is in under-utilization or over-utilization state in a dynamic environment. A good task scheduling technique is always required for the dynamic allocation of the task to avoid such a situation. Through this paper we are going to present the Genetic Algorithm based task scheduling technique, which will distribute the load effectively among the virtual machine so that the overall response time (QoS) should be minimal. A comparison of this Genetic Algorithm based task scheduling technique is performed on CloudSim simulator which shows that, this will outperform the existing techniques like Greedy based, First — Come first — Serve (FCFS) techniques.

We describe a new approach to speech recognition, in which all Hidden Markov Model (HMM) states share the same Gaussian Mixture Model (GMM) structure with the same number of Gaussians in each state. The model is defined by vectors associated with each state with a dimension of, say, 50, together with a global mapping from this vector space to the space of parameters of the GMM. This model appears to give better results than a conventional model, and the extra structure offers many new opportunities for modeling innovations while maintaining compatibility with most standard techniques.