Search

In this letter, we propose a continual learning approach for a set of sequential scene classification tasks, where each task contains a group of land-cover classes. Our aim is to learn new tasks in a continual way without significantly degrading the performances of the old ones, due to the tricky catastrophic forgetting problem inherent to neural networks. To this end, we propose a neural architecture composed of two trainable modules. The first module learns its weights by discriminating between the land-cover classes within the new task while keeping trace of the old ones. On the other side, the second module tries to maximize the separation between the tasks by learning on task-prototypes stored in a linear memory (one prototype per task). The experimental results on two scene data sets (Merced and Optimal31) confirm the promising capability of the proposed method.

Scene classification is a crucial research problem in remote sensing (RS) that has attracted many researchers recently. It has many challenges due to multiple issues, such as: the complexity of remote sensing scenes, the classes overlapping (as a scene may contain objects that belong to foreign classes), and the difficulty of gaining sufficient labeled scenes. Deep learning (DL) solutions and in particular convolutional neural networks (CNN) are now state-of-the-art solution in RS scene classification; however, CNN models need huge amounts of annotated data, which can be costly and time-consuming. On the other hand, it is relatively easy to acquire large amounts of unlabeled images. Recently, Self-Supervised Learning (SSL) is proposed as a method that can learn from unlabeled images, potentially reducing the need for labeling. In this work, we propose a deep SSL method, called RS-FewShotSSL, for RS scene classification under the few shot scenario when we only have a few (less than 20) labeled scenes per class. Under this scenario, typical DL solutions that fine-tune CNN models, pre-trained on the ImageNet dataset, fail dramatically. In the SSL paradigm, a DL model is pre-trained from scratch during the pretext task using the large amounts of unlabeled scenes. Then, during the main or the so-called downstream task, the model is fine-tuned on the labeled scenes. Our proposed RS-FewShotSSL solution is composed of an online network and a target network both using the EfficientNet-B3 CNN model as a feature encoder backbone. During the pretext task, RS-FewShotSSL learns discriminative features from the unlabeled images using cross-view contrastive learning. Different views are generated from each image using geometric transformations and passed to the online and target networks. Then, the whole model is optimized by minimizing the cross-view distance between the online and target networks. To address the problem of limited computation resources available to us, our proposed method uses a novel DL architecture that can be trained using both high-resolution and low-resolution images. During the pretext task, RS-FewShotSSL is trained using low-resolution images, thereby, allowing for larger batch sizes which significantly boosts the performance of the proposed pipeline on the task of RS classification. In the downstream task, the target network is discarded, and the online network is fine-tuned using the few labeled shots or scenes. Here, we use smaller batches of both high-resolution and low-resolution images. This architecture allows RS-FewshotSSL to benefit from both large batch sizes and full image sizes, thereby learning from the large amounts of unlabeled data in an effective way. We tested RS-FewShotSSL on three RS public datasets, and it demonstrated a significant improvement compared to other state-of-the-art methods such as: SimCLR, MoCo, BYOL and IDSSL.

In this work, we introduce a neural architecture to unsupervised domain from multiple source domains. This architecture uses an EfficientNet as a feature extractor coupled with a set of Softmax classifiers equal to the number of source domains followed by an opportune fusion layer. To reduce the domain discrepancy between each source and target domain, we adopt a Minmax entropy approach that is based on the idea of optimizing in an adversarial manner the conditional entropy of the target samples with respect to each source classifier and minimizes it with respect to the feature extractor. As for the fusion module, we propose a weighted average fusion layer with learnable weights for aggregating the outputs of the different Softmax classifiers. Experiments on a multisource data set composed of images acquired by manned and unmanned aerial vehicles (MAVs/UAVs) over different locations are reported and discussed.

Scene classification is a highly useful task in Remote Sensing (RS) applications. Many efforts have been made to improve the accuracy of RS scene classification. Scene classification is a challenging problem, especially for large datasets with tens of thousands of images with a large number of classes and taken under different circumstances. One problem that is observed in scene classification is the fact that for a given scene, only one part of it indicates which class it belongs to, whereas the other parts are either irrelevant or they actually tend to belong to another class. To address this issue, this paper proposes a deep attention Convolutional Neural Network (CNN) for scene classification in remote sensing. CNN models use successive convolutional layers to learn feature maps from larger and larger regions (or receptive fields) of the scene. The attention mechanism computes a new feature map as a weighted average of these original feature maps. In particular, we propose a solution, named EfficientNet-B3-Attn-2, based on the pre-trained EfficientNet-B3 CNN enhanced with an attention mechanism. A dedicated branch is added to layer 262 of the network, to compute the required weights. These weights are learned automatically by training the whole CNN model end-to-end using the backpropagation algorithm. In this way, the network learns to emphasize important regions of the scene and suppress the regions that are irrelevant to the classification. We tested the proposed EfficientNet-B3-Attn-2 on six popular remote sensing datasets, namely UC Merced, KSA, OPTIMAL-31, RSSCN7, WHU-RS19, and AID datasets, showing its strong capabilities in classifying RS scenes.

Advances in technology can provide a lot of support for visually impaired (VI) persons. In particular, computer vision and machine learning can provide solutions for object detection and recognition. In this work, we propose a multi-label image classification solution for assisting a VI person in recognizing the presence of multiple objects in a scene. The solution is based on the fusion of two deep CNN models using the induced ordered weighted averaging (OWA) approach. Namely, in this work, we fuse the outputs of two pre-trained CNN models, VGG16 and SqueezeNet. To use the induced OWA approach, we need to estimate a confidence measure in the outputs of the two CNN base models. To this end, we propose the residual error between the predicted output and the true output as a measure of confidence. We estimate this residual error using another dedicated CNN model that is trained on the residual errors computed from the main CNN models. Then, the OAW technique uses these estimated residual errors as confidence measures and fuses the decisions of the two main CNN models. When tested on four image datasets of indoor environments from two separate locations, the proposed novel method improves the detection accuracy compared to both base CNN models. The results are also significantly better than state-of-the-art methods reported in the literature.

We present a stochastic deep collocation method (DCM) based on neural architecture search (NAS) and transfer learning for heterogeneous porous media. We first carry out a sensitivity analysis to determine the key hyper-parameters of the network to reduce the search space and subsequently employ hyper-parameter optimization to finally obtain the parameter values. The presented NAS based DCM also saves the weights and biases of the most favorable architectures, which is then used in the fine-tuning process. We also employ transfer learning techniques to drastically reduce the computational cost. The presented DCM is then applied to the stochastic analysis of heterogeneous porous material. Therefore, a three dimensional stochastic flow model is built providing a benchmark to the simulation of groundwater flow in highly heterogeneous aquifers. The performance of the presented NAS based DCM is verified in different dimensions using the method of manufactured solutions. We show that it significantly outperforms finite difference methods in both accuracy and computational cost.

The electrocardiogram (ECG) signal is shown to be promising as a biometric. To this end, it has been demonstrated that the analysis of ECG signals can be considered as a good solution for increasing the biometric security levels. This can be mainly due to its inherent robustness against presentation attacks. In this work, we present a deep contrastive learning-based system for ECG biometric identification. The proposed system consists of three blocks: a feature extraction backbone based on short time Fourier transform (STFT), a contrastive learning network, and a classification network. We evaluated the proposed system on the Heartprint dataset, a new ECG biometrics multi-session dataset. The experimental analysis shows promising capabilities of the proposed method. In particular, it yields an average top1 accuracy of 98.02% on a new dataset built by gathering 1539 ECG records from 199 subjects collected in multiple sessions with an average interval between sessions of 47 days.

Saudi Arabia has implemented its ambitious and comprehensive national strategy, i.e., Saudi Vision 2030, to achieve major economic, social, and environmental objectives. The main aim of this paper is to study the Granger causality relationships between economic growth, environmental degradation, and urbanization in Saudi Arabia over the period from 1985 to 2019. At first, Augmented Dicky-Fuller (ADF) and Phillips-Perron (PP) tests were applied in order to check the stationarity of the panel time-series data. Since the data were of mixed order of integration I(0) and I(1), the Autoregressive Distributed Lag (ARDL) framework was employed to perform the statistical analysis. Then, the short- and long-run relationships were evaluated using the bounds test for cointegration applied on the Error Correction Models (ECMs) for GDP, CO2 emissions, and urbanization as the dependent variables. Furthermore, the direction and significance of causality were estimated in the ARDL/ECM framework. In addition, the Environmental Kuznets Curve (EKC) hypothesis was examined for the sample data. To assess the generalization capability of the findings in this study, robustness and diagnostic tests were applied. In the long-run, the empirical findings indicate that 1% increase in economic growth Granger caused 0.15% increase in CO2 emissions and 0.006% decrease in urbanization. Whereas 1% increase in urbanization Granger caused 2.5% increase in the economic growth. In the short-run, a unidirectional causal relationship existed from economic growth to both CO2 emissions and urbanization with 1% increase in GDP Granger caused 0.3% and 0.004% increases in CO2 emissions and urbanization, respectively. Finally, policy recommendations were presented in light of the Saudi Vision 2030.

The steady spread of the 2019 Coronavirus disease has brought about human and economic losses, imposing a new lifestyle across the world. On this point, medical imaging tests such as computed tomography (CT) and X-ray have demonstrated a sound screening potential. Deep learning methodologies have evidenced superior image analysis capabilities with respect to prior handcrafted counterparts. In this paper, we propose a novel deep learning framework for Coronavirus detection using CT and X-ray images. In particular, a Vision Transformer architecture is adopted as a backbone in the proposed network, in which a Siamese encoder is utilized. The latter is composed of two branches: one for processing the original image and another for processing an augmented view of the original image. The input images are divided into patches and fed through the encoder. The proposed framework is evaluated on public CT and X-ray datasets. The proposed system confirms its superiority over state-of-the-art methods on CT and X-ray data in terms of accuracy, precision, recall, specificity, and F1 score. Furthermore, the proposed system also exhibits good robustness when a small portion of training data is allocated.

We present a deep learning method to investigate the effect of flexoelectricity in nanostructures. For this purpose, deep neural network (DNN) algorithm is employed to map the relation between the inputs and the material response of interest. The DNN model is trained and tested making use of database that has been established by solving the governing equations of flexoelectricity using a NURBS-based IGA formulation at design points in the full probability space of the input parameters. Firstly, pure flexoelectric cantilever nanobeam is investigated under mechanical and electrical loading conditions. Then, structures of composite system constituted by two non-piezoelectric material phases are addressed in order to find the optimized topology with respect to the energy conversion factor. The results show promising capabilities of the proposed method, in terms of accuracy and computational efficiency. The deep learning method we used have produced superior optimal designs compared to the numerical methods. The findings of this study will be of profound interest to researcher involved further in the optimization and design of flexoelectric structures.

We discuss how the Dempster-Shafer belief structure provides a framework for modeling an uncertain value x ˜ from some domain X. We note how it involves a two-step process: the random determination of one focal element (set) guided by a probability distribution and then the selection of x ˜ from this focal element in some unspecified manner. We generalize this framework by allowing the selection of the focal element to be determined by a random experiment guided by a fuzzy measure. In either case the anticipation that x ˜ lies in some subset E is interval-valued, [Bel(E), Pl(E)]. We next look at database retrieval and turn to issue of determining if a database entity with an uncertain attribute value satisfies a desired value. Here we model our uncertain attribute value as x ˜ and our desired value as a subset E. In this case the degree of satisfaction of the query E by the entity is [Bel(E), Pl(E)]. In order to compare these interval-valued satisfactions we use the Golden rule representative value to turn the intervals into scalars. We describe an application involving retrieval from a uncertain database.

Experimentally fabricated two-dimensional materials have lately evoked significant attention in the nanodevice fabrication industry. In this manuscript, we study the mechanical response of crystalline boron-carbide with BC3 stoichiometry, which is a novel two-dimensional (2D) graphene-like material. Its excellent electrical, thermal and mechanical properties make it an exceptional candidate for a wide range of applications. However, different type of defects created during the production process or device assembly may deteriorate its remarkable mechanical properties. Hence, the purpose of this study is to investigate the mechanical properties of pristine and defective BC3 nanosheets through classical molecular dynamics (MD) simulations. Therefore, we study for the first time the influence of several crack lengths and notch diameters on the mechanical response at different temperatures under uniaxial tensile loading. Our results indicate that larger cracks and notches decrease the strength of 2D graphene-like BC3 nanosheets. Additionally, it was revealed that a temperature increase induces a weakening effect on the tensile strength of BC3 monolayer. Our MD results not only highlight the outstanding mechanical properties of graphene-like BC3, but also reveal its advantages regarding its thermo-mechanical properties, which are critical for the design of nanodevices.

Two-dimensional (2D) materials have attracted the interests of various research communities in material science due to their unique properties and broad application prospects. The experimental advances achieved during the last decade facilitate the fabrication of novel 2D structures with a wide range of applications in nanodevices. A recent experimental study (Nat. Commun. v.7, 13774, 2016) provided a synthesis route and confirmed the structural and electronic properties of novel 2D layered RuCl3 nanosheets. These materials have displayed Kitaev physics. Owing to its stable atomic lattice and very appealing magnetic properties, the single layer RuCl3 is an important material for producing chemical catalysts with applications in nanoelectronics. Motivated by recent experimental advances, we conducted first-principles calculations to study the dynamic behaviour and mechanical characteristics of pristine RuCl3 and RuBr3 in their single-layer form. We performed spin-polarized density functional theory calculations of specimen subjected to uniaxial tensile loading to predict the mechanical/failure properties of these novel 2D materials. Analyzing the phonon dispersions confirmed the dynamic stability of the stress-free atomic lattices. Our density functional theory (DFT) results also reveal important mechanical properties of RuCl3/RuBr3 as a class of super-stretchable 2D materials which are appealing for nanodevices.

A method for shape optimization using Bezier triangles is introduced. The proposed procedure takes as input a CAD-compatible boundary representation of the domain and outputs an optimal design while maintaining an exact geometry representation at each iteration. The use of a triangular discretization allows the modeling of complex geometric domains, including voids, using a single patch. Some topology changes, such as those resulting from merging boundaries, can also be easily considered. An automatic mesh generator based on a quadtree construction is used to create the mesh. A gradient-based optimization algorithm (the method of moving asymptotes) is employed together with a sensitivity propagation procedure. We apply the method to some standard benchmark problems commonly considered in the literature and show that the proposed method converges to an optimal shape in only a few iterations.

In this paper, we present a new approach to identify all material parameters of flexoelectric materials based on electrical impedance curves. This approach combines an Isogeometric Analysis (IGA) formulation with a gradient-based optimization algorithm using the Method of Moving Asymptotes (MMA). The IGA formulation allows for efficient modeling of flexoelectricity taking advantage of the higher order continuity of IGA. The proposed methodology starts with determining preliminary real parts based on resonant modes in order to avoid local minima which gives the numerical impedance curves close to the experimental impedance curve. The results in the preliminary step are used as initial parameters of the refinement step to simultaneously determine both real and imaginary part by minimizing the difference between pseudo-experimental and numerical impedance curve. Some numerical examples are illustrated to show the good agreement between the numerical and pseudo-experimental impedance curves.

The electrocardiogram (ECG) signal produced by the human heart is an emerging biometric modality that can play an important role in the future generation’s identity recognition with the support of machine learning techniques. One of the major obstacles in the progress of this modality is the lack of public datasets with a long interval between sessions of data acquisition to verify the uniqueness and permanence of the biometric signature of the heart of a subject. To address this issue, we put forward Heartprint, a large biometric database of multisession ECG signals comprising 1539 records captured from the fingers of 199 healthy subjects. The capturing time for each record was 15 s, and recordings were made in resting and reading conditions. They were collected in multiple sessions over ten years, and the average interval between first session (S1) and third session (S3L) was 1572.2 days. The dataset also covers several demographic classes such as genders, ethnicities, and age groups. The combination of raw ECG signals and demographic information turns the Heartprint dataset, which is made publicly available online, into a valuable resource for the development and evaluation of biometric recognition algorithms.

With the rapid growth of fingerprint-based biometric systems, it is essential to ensure the security and reliability of the deployed algorithms. Indeed, the security vulnerability of these systems has been widely recognized. Thus, it is critical to enhance the generalization ability of fingerprint presentation attack detection (PAD) cross-sensor and cross-material settings. In this work, we propose a novel solution for addressing the case of a single source domain (sensor) with large labeled real/fake fingerprint images and multiple target domains (sensors) with only few real images obtained from different sensors. Our aim is to build a model that leverages the limited sample issues in all target domains by transferring knowledge from the source domain. To this end, we train a unified generative adversarial network (UGAN) for multidomain conversion to learn several mappings between all domains. This allows us to generate additional synthetic images for the target domains from the source domain to reduce the distribution shift between fingerprint representations. Then, we train a scale compound network (EfficientNetV2) coupled with multiple head classifiers (one classifier for each domain) using the source domain and the translated images. The outputs of these classifiers are then aggregated using an additional fusion layer with learnable weights. In the experiments, we validate the proposed methodology on the public LivDet2015 dataset. The experimental results show that the proposed method improves the average classification accuracy over twelve classification scenarios from 67.80 to 80.44% after adaptation.

Numerous studies addressed the impacts of social development and economic growth on the environment. This paper presents a study about the inclusive impact of social and economic factors on the environment by analyzing the association between carbon dioxide (CO2) emissions and two socioeconomic indicators, namely, Human Development Index (HDI) and Legatum Prosperity Index (LPI), under the Environmental Kuznets Curve (EKC) framework. To this end, we developed a two-stage methodology. At first, a multivariate model was constructed that accurately explains CO2 emissions by selecting the appropriate set of control variables based on model quality statistics. The control variables include GDP per capita, urbanization, fossil fuel consumption, and trade openness. Then, quantile regression was used to empirically analyze the inclusive relationship between CO2 emissions and the socioeconomic indicators, which revealed many interesting results. First, decreasing CO2 emissions was coupled with inclusive socioeconomic development. Both LPI and HDI had a negative marginal relationship with CO2 emissions at quantiles from 0.2 to 1. Second, the EKC hypothesis was valid for G20 countries during the study period with an inflection point around quantile 0.15. Third, the fossil fuel consumption had a significant positive relation with CO2 emissions, whereas urbanization and trade openness had a negative relation during the study period. Finally, this study empirically indicates that effective policies and policy coordination on broad social, living, and economic dimensions can lead to reductions in CO2 emissions while preserving inclusive growth.

Fingerprint-based biometric systems have grown rapidly as they are used for various applications including mobile payments, international border security, and financial transactions. The widespread nature of these systems renders them vulnerable to presentation attacks. Hence, improving the generalization ability of fingerprint presentation attack detection (PAD) in cross-sensor and cross-material setting is of primary importance. In this work, we propose a solution based on a transformers and generative adversarial networks (GANs). Our aim is to reduce the distribution shift between fingerprint representations coming from multiple target sensors. In the experiments, we validate the proposed methodology on the public LivDet2015 dataset provided by the liveness detection competition. The experimental results show that the proposed architecture yields an increase in average classification accuracy from 68.52% up to 83.12% after adaptation.

Glaucoma is one of the disorders that infects the retinal. All people are exposed to infection by Glaucoma, but age people most commonly affect them and lead to loss of vision. Unfortunately, there is no Glaucoma medication yet, but the good news is, early detection of it prevents further vision loss or blindness. The traditional diagnose of Glaucoma faced many challenges like a long time, less of ophthalmologists in the remote area, and difficulty detection Glaucoma in the early stage of it. Therefore, clinical diagnosis has been combined with computer vision techniques. In this paper, we suggest a deep learning method based on Cup-to-disc ratio measures for the detection of Glaucoma. We used Encoder-Decoder with Atrous Convolution and a Self-attention mechanism, which allows modeling a long-range dependency across image regions. The experimental results of this method are proved in the REFUGE dataset.

Glaucoma is one of the main retinal diseases. Glaucoma affects older people more often, and it can lead to vision loss. Until now there is no medicament for Glaucoma, but early detection is important, wherein it can limit the increase of vision loss or blindness. In this paper, we propose a deep learning approach based on two steps for Glaucoma detection in retinal fundus images. In the first step, we use a faster region proposal neural network (RCNN) to detect the optical disc (OD). Then in a second step, we train a regression network to estimate the cup-to-disc ratio (CDR) by analyzing reign around the detected OD. Experimental results of this method are demonstrated on the MESSIDOR and Magrabi datasets.

Most of the existing domain adaptation (DA) methods proposed in the context of remote sensing imagery assume the presence of the same land-cover classes in the source and target domains. Yet, this assumption is not always realistic in practice as the target domain may contain additional classes unknown to the source leading to the so-called open set DA. Under this challenging setting, the problem turns to reducing the distribution discrepancy between the shared classes in both domains besides the detection of the unknown class samples in the target domain. To deal with the openset problem, we propose an approach based on adversarial learning and pareto-based ranking. In particular, the method leverages the distribution discrepancy between the source and target domains using min-max entropy optimization. During the alignment process, it identifies candidate samples of the unknown class from the target domain through a pareto-based ranking scheme that uses ambiguity criteria based on entropy and the distance to source class prototype. Promising results using two cross-domain datasets that consist of very high resolution and extremely high resolution images, show the effectiveness of the proposed method.

In this paper, we propose a method for localizing the optic nerve head and segmenting the optic disc/cup in retinal fundus images. The approach is based on a simple two-stage Mask-RCNN compared to sophisticated methods that represent the state-of-the-art in the literature. In the first stage, we detect and crop around the optic nerve head then feed the cropped image as input for the second stage. The second stage network is trained using a weighted loss to produce the final segmentation. To further improve the detection in the first stage, we propose a new fine-tuning strategy by combining the cropping output of the first stage with the original training image to train a new detection network using different scales for the region proposal network anchors. We evaluate the method on Retinal Fundus Images for Glaucoma Analysis (REFUGE), Magrabi, and MESSIDOR datasets. We used the REFUGE training subset to train the models in the proposed method. Our method achieved 0.0430 mean absolute error in the vertical cup-to-disc ratio (MAE vCDR) on the REFUGE test set compared to 0.0414 obtained using complex and multiple ensemble networks methods. The models trained with the proposed method transfer well to datasets outside REFUGE, achieving a MAE vCDR of 0.0785 and 0.077 on MESSIDOR and Magrabi datasets, respectively, without being retrained. In terms of detection accuracy, the proposed new fine-tuning strategy improved the detection rate from 96.7% to 98.04% on MESSIDOR and from 93.6% to 100% on Magrabi datasets compared to the reported detection rates in the literature.

In this paper, we propose a novel end-to-end learnable architecture based on Dense Convolutional Networks (DCN) for the classification of electrocardiogram (ECG) signals. This architecture is based on two main modules: the first is a generative module and the second is a discriminative one. The task of the generative module is to convert the one dimensional ECG signal into an image by means of fully connected, up-sampling, and convolution layers. The discriminative module takes as input the generated image and carries out feature learning and classification. To handle the data imbalance problem characterizing the ECG data, we propose to use the focal loss (FL) that is based on the idea of reshaping the standard cross-entropy loss such that it reduces the loss assigned to well-classified ECG beats. In the experiments, we validate the method using the well-known MIT-BIH arrhythmia database in four different scenarios, using four classes in the first scenario, five in the second and 12 in the third. Finally, supraventricular versus the other three and ventricular versus the other three from the scenario with four classes are used as the fourth scenario. The results obtained show that the method proposed here achieves a significant accuracy improvement over all previous state-of-the-art methods.

The multi-label classification problem in Unmanned Aerial Vehicle (UAV) images is particularly challenging compared to single-label classification due to its combinatorial nature. To tackle this issue, we propose in this paper a deep learning approach based on encoder-decoder neural network architecture with channel and spatial attention mechanisms. Specifically, the encoder module which is based on a pre-trained convolutional neural network (CNN) has the task to transform the input image to a set of feature maps using an opportune feature combination. To improve the feature representation further, this module incorporates a squeeze excitation (SE) layer for modelling the interdependencies between the channels of the feature maps. The decoder module which is based on a long short terms memory (LSTM) network has the task of generating, in a sequential way, the classes present in the image. At each time step, it predicts the next class-label by aligning its hidden state to the corresponding region in the image by means of an adaptive spatial attention mechanism. The experiments carried out on two UAV datasets with a spatial resolution of 2-cm show that our method is promising in predicting the labels present in the image while attending the relevant objects in the image. Additionally, it is able to provide better classification results compared to state-of-the-art methods.

This manuscript introduces a computational approach to micro-damage problems using deep learning for the prediction of loading deflection curves. The location of applied forces, dimensions of the specimen and material parameters are used as inputs of the process. The micro-damage is modelled with a gradient-enhanced damage model which ensures the well-posedness of the boundary value and yields mesh-independent results in computational methods such as FEM. We employ the Adam optimizer and Rectified linear unit activation function for training processes and research into the deep neural network architecture. The performance of our approach is demonstrated through some numerical examples including the three-point bending specimen, shear bending on L-shaped specimen and different failure mechanisms.

Electromechanical coupling devices have been playing an indispensable role in modern engineering. Particularly, flexoelectricity, an electromechanical coupling effect that involves strain gradients, has shown promising potential for future miniaturized electromechanical coupling devices. Therefore, simulation of flexoelectricity is necessary and inevitable. In this paper, we provide an overview of numerical procedures on modeling flexoelectricity. Specifically, we summarize a generalized formulation including the electrostatic stress tensor, which can be simplified to retrieve other formulations from the literature. We further show the weak and discretization forms of the boundary value problem for different numerical methods, including isogeometric analysis and mixed FEM. Several benchmark problems are presented to demonstrate the numerical implementation. The source code for the implementation can be utilized to analyze and develop more complex flexoelectric nano-devices.

Multi-label classification problem aims to assign multiple class labels to the remote sensing image under analysis, which is more challenging compared to single-label classification. To this end, we propose a neural model based on multiple loss functions. The first loss seeks to increase the similarity between the image with its corresponding labels using a similarity layer. The second one is related to label discrimination, and it is achieved using a modified softmax layer suitable for multi-label classification. The third loss aims to detect automatically the number of labels present in the image through a regression layer. Experimental results on the well known Merced data are reported and discussed.

Scene classification has become an important research topic in remote sensing (RS) field. Typical solution relies on labeling a large enough set of the RS scenes manually using expert opinion if needed, then training the algorithm on this set to learn how to correctly classify other new scenes. The best performance deep learning models required a large labeled dataset for training. Accordingly, there is great need to develop intelligent machine learning algorithm that can learn to classify RS datasets containing new unseen classes from few labeled samples only. This problem is known as few-shot machine learning. In this work we develop a deep few-shot learning method for the classification of RS scenes. The proposed method is based on prototypical deep neural networks combined with SqueezeNet pre-trained CNN for image embedding. In this paper, we report preliminary results using the two RS scene datasets UC Merced and optimal31.

In this paper, an approximate solution based on the continuum mechanics is presented to obtain the mechanical response of freestanding single and two layers molybdenum disulfide (MoS2) films. Both circular and rectangular geometries as well as point load and uniform pressure are investigated. The partial differential equilibrium equations are established based on the non-linear Foppl membrane theory for both circular and rectangular films. The Galerkin method is then employed to solve the non-linear partial differential equilibrium equations in case of circular films. Also, the principle of virtual work affords another means of obtaining an approximate solution for rectangular film. Theoretical predictions for the width of membrane regions enable us to estimate the analytical deflection and stress distributions. Finally, the obtained results are compared to the available experimental and theoretical results and a good agreement is seen. It can be concluded that the continuum approach can be used to study the behavior of the MoS2 films.

Graphene is a promising material for nanopore-sequencing DNA technology. In the current study, we have utilized molecular dynamic simulation in order to fabricate nanopores in the four divergent locations with different properties in a single-layer graphene nanosheet by using clusters bombardment. Ten different kinetic energies have been applied to three different diameters of SiC, Si and diamond clusters to fabricate nanopores. Image processing technology has also been applied to compute the exact area of regular and irregular drilled nanopores. The obtained results suggest that the desired size and qualities of nanopore can be achieved by controlling the type, diameter and the energy of the clusters. We have observed that the average area of nanopores increases by rising the kinetic energy in the most of cases. Moreover, the properties of the incident location can highly affect the area size and quality of the nanopores. The largest area size of the nanopores has been obtained when the incident location is placed in the center of the grains, while the smallest area of nanopores have been observed when the incident point is placed on the grain boundaries junction. Among all three types of clusters, the impact of the diamond cluster with diameter of 2 nm fabricates the most suitable nanopores. Therefore, we have used the diamond cluster to investigate the effect of straining the nanosheet on the topography of nanopores. We applied 3% and 5% of tensile and compressive strain to the graphene nanosheet. Under tensile strains we found that increasing the external tensile strain on the nanosheet fabricates larger nanopores with smoother edges. On the other hand, applying external compressive strains leads to nanopores with more irregular topography.

In this paper, sensitivity analysis has been applied to identify the key input parameters influencing the energy conversion factor (ECF) of flexoelectric materials. The governing equations of flexoelectricity are modeled by a NURBS-based IGA formulation exploiting their higher order continuity and hence avoiding a complex mixed formulation. The examined input parameters include model and material properties, and the sampling has been obtained using the latin hypercube sampling (LHS) method in the probability space. The sensitivity of the model output to each of the input parameters at different aspect ratios of the beam is quantified by three various common methods, i.e. Morris One-At-a-Time (MOAT), PCE-Sobol’, and Extended Fourier amplitude sensitivity test (EFAST). The numerical results indicate that the flexoelectric constants are the most dominant factors influencing the uncertainties in the energy conversion factor, in particular the transversal flexoelectric coefficient ( h 12 ) . Moreover, the model parameters also show considerable interaction effects of the material properties.

The purpose of this study is to investigate the mechanical properties of a new two-dimensional graphene like material, crystalline carbon nitride with the stoichiometry of C3N. The extraordinary properties of C3N make it an outstanding candidate for a wide variety of applications. In this work, the mechanical properties of C3N nanosheets have been studied not only in the defect-free form, but also with critical defects such as line cracks and notches using molecular dynamics simulations. Different crack lengths and notch diameters were considered to predict the mechanical response at different temperatures under the uniaxial tensile loading. Our simulation results show that larger cracks and notches reduce the strength of the nanosheets. Moreover, it was shown the temperature rise has a weakening effect on the tensile strength of C3N. Our study can provide useful information with respect to the thermo-mechanical properties of pristine and defective graphene like C3N 2D material.

In the present paper, molecular dynamic simulations have been conducted to investigate the nanopores creation on 10% of boron and nitrogen doped polycrystalline graphene by silicon and diamond nanoclusters . Two types of nanoclusters based on silicon and diamond are used to investigate their effect for the fabrication of nanopores. Therefore, three different diameter sizes of the clusters with five kinetic energies of 10, 50, 100, 300 and 500 eV/atom at four different locations in boron or nitrogen doped polycrystalline graphene nanosheets have been perused. We also study the effect of 3% and 6% of boron doped polycrystalline graphene with the best outcome from 10% of doping. Our results reveal that the diamond cluster with diameter of 2 and 2.5 nm fabricates the largest nanopore areas on boron and nitrogen doped polycrystalline graphene, respectively. Furthermore, the kinetic energies of 10 and 50 eV/atom can not fabricate nanopores in some cases for silicon and diamond clusters on boron doped polycrystalline graphene nanosheets. On the other hand, silicon and diamond clusters fabricate nanopores for all locations and all tested energies on nitrogen doped polycrystalline graphene. The area sizes of nanopores fabricated by silicon and diamond clusters with diameter of 2 and 2.5 nm are close to the actual area size of the related clusters for the kinetic energy of 300 eV/atom in all locations on boron doped polycrystalline graphene. The maximum area and the average maximum area of nanopores are fabricated by the kinetic energy of 500 eV/atom inside the grain boundary at the center of the nanosheet and in the corner of nanosheet with diameters of 2 and 3 nm for silicon and diamond clusters on boron and nitrogen doped polycrystalline graphene.

The accurate reconstruction of areas obscured by clouds is among the most challenging topics for the remote sensing community since a significant percentage of images archived throughout the world are affected by cloud covers which make them not fully exploitable. The purpose of this paper is to propose new methods to recover missing data in multispectral images due to the presence of clouds by relying on a formulation based on an autoencoder (AE) neural network. We suppose that clouds are opaque and their detection is performed by dedicated algorithms. The AE in our methods aims at modeling the relationship between a given cloud-free image (source image) and a cloud-contaminated image (target image). In particular, two strategies are developed: the first one performs the mapping at a pixel level while the second one at a patch level to take profit from spatial contextual information. Moreover, in order to fix the problem of the hidden layer size, a new solution combining the minimum descriptive length criterion and a Pareto-like selection procedure is introduced. The results of experiments conducted on three different data sets are reported and discussed together with a comparison with reference techniques.

We discuss the role that the Choquet integral plays in the aggregation of criteria satisfaction in multi-criteria decision functions. We show how the choice of the associated measure allows for the formulation of many types of multi-criteria decision functions. We note that the need for an ordering of the criteria satisfactions causes difficulties in situations in which there exists a probabilistic type of uncertainty in the knowledge of the criteria satisfactions. We discuss an approach, called the probabilistic exceedance method, for allowing the aggregation of probabilistically satisfied criteria.

In this work, we present a Parametric Deep Energy Method (P-DEM) for elasticity problems accounting for strain gradient effects. The approach is based on physics-informed neural networks (PINNs) for the solution of the underlying potential energy. Therefore, a cost function related to the potential energy is subsequently minimized. P-DEM does not need any classical discretization and requires only a definition of the potential energy, which simplifies the implementation. Instead of training the model in the physical space, we define a parametric/reference space similar to isoparametric finite elements, which is in our example a unit square. The inputs are naturally normalized preventing the vanishing gradient problem and leading to much faster convergence compared to the original DEM. Forward–backward mapping is established by means of NURBS basis functions. Another advantage of this approach is that Gauss quadrature can be employed to approximate the total potential energy , which is the loss function calculated in the parametric domain . Backpropagation available in PyTorch with automatic differentiation is performed to calculate the gradients of the loss function with respect to the weights and biases. Once the network is trained, a numerical solution can be obtained in the reference domain and then is mapped back to the physical domain. The performance of the method is demonstrated through various numerical benchmark problems in elasticity and compared to analytical solutions. We also consider strain gradient elasticity, which poses challenges to conventional finite elements due to the requirement for C 1 continuity.

Most existing Electrocardiogram (ECG) classification methods assume that all arrhythmia classes are known during the training phase. In this paper, the problem of learning several successive tasks is addressed, where, in each new task, there are new arrhythmia classes to learn. Unfortunately, in machine learning it is known that when a model is retrained onto a new task, the machine tends to forget the old task. This is known in machine learning, as 'the catastrophic forgetting phenomenon'. To this end, a learn-without-forgetting (LwF) approach to solve this problem is proposed. This novel deep LwF method for ECG heartbeat classification is the first work of its kind in the field. This proposed LwF approach consists of a deep learning architecture that includes the following important aspects: feature extraction module, classification layers for each learned task, memory module to store one prototype for each task, and a task selection module able to identify the most suitable task for each input sample. The feature extraction module constitutes another contribution of this work. It starts with a set of deep layers that convert an ECG heartbeat signal into an image, then the pre-trained DenseNet169 CNN takes the obtained image and extracts rich and powerful features that are effective inputs for the classifications layers of the model. Whenever a new task is to be learned, the network expands with a new classification layer having a Softmax activation function. The newly added layer is responsible for learning the classes of the new task. When the network is trained for the new task, the shared layers, as well as the output layers of the old tasks, are also fine-tuned using pseudo labels. This helps in retaining knowledge of old tasks. Finally, the task selector stores feature prototypes for each task, and using a distance matching network, is trained to select which task is more suitable to classify a new test sample. The whole network uses end-to-end learning to optimize one loss functions, which is a weighted combination of the loss functions of the different network modules. The proposed model was tested on three common ECG datasets, namely the MIT-BIH, INCART, and SVDB datasets. The results obtained demonstrate the success of the proposed method in learning, without forgetting, successive ECG heartbeat classification tasks.