Search

In the original publication [...]

This study introduces the Divergent Selective Focused Multi-heads Self-Attention Network (DSFMANet), an innovative deep learning model devised to automatically predict Hamilton Depression Rating Scale-17 (HAMD-17) scores in patients with depression. This model introduces a multi-branch structure for sub-bands and artificially configures attention focus factors on various branches, resulting in distinct attention distributions for different sub-bands. Experimental results demonstrate that when DSFMANet processes sub-band data, its performance surpasses current benchmarks in key metrics such as mean square error (MSE), mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R2). This success is particularly evident in terms of MSE and MAE, where the performance of sub-band data is significantly superior, highlighting the model's potential in accurately predicting HAMD-17 scores. Concurrently, the experiment also compared the model's performance with sub-band and full-band data, affirming the superiority of the selective focus attention mechanism in electroencephalography (EEG) signal processing. DSFMANet, when utilizing sub-band data, exhibits higher data processing efficiency and reduced model complexity. The findings of this study underscore the significance of employing deep learning models based on sub-band analysis in depression diagnosis. The DSFMANet model not only effectively enhances the accuracy of depression diagnosis but also offers valuable research directions for similar applications in the future. This deep learning-based automated approach can effectively ascertain the HAMD-17 scores of patients with depression, thus offering more accurate and reliable support for clinical decision-making.

This study tackles the significant challenge of generating low-cost intrusion detection datasets for Internet of Things (IoT) camera devices, particularly for financially limited organizations. Traditional datasets often depend on costly cameras, posing accessibility issues. Addressing this, a new dataset was developed, tailored for low-cost IoT devices, focusing on essential features. The research employed an Entry/Exit IoT Network at CKT-UTAS, Navrongo, a Ghanaian University, showcasing a feasible model for similar organizations. The study gathered location and other vital features from low-cost cameras and a standard dataset. Using the XGBoost machine learning algorithm, the effectiveness of this approach for cybersecurity enhancement was demonstrated. The implementation included a model-agnostic eXplainable AI (XAI) technique, employing Shapley Additive Explanations (SHAP) values to interpret the XGBoost model's predictions. This highlighted the significance of cost-effective features like Flow Duration, Total Forward Packets, and Total Length Forward Packet, in addition to location data. These features were crucial for intrusion detection using the new IoT dataset. Training a deep-learning model with only these features maintained comparable accuracy to using the full dataset, validating the practicality and efficiency of the approach in real-world scenarios.

Graph Neural Networks (GNNs) have gained popularity in image matching methods, proving useful for various computer vision tasks like Structure from Motion (SfM) and 3D reconstruction. A well-known example is SuperGlue. Lightweight variants, such as LightGlue, have been developed with a focus on stacking fewer GNN layers compared to SuperGlue. This paper proposes the h-GNN, a lightweight image matching model, with improvements in the two processing modules, the GNN and matching modules. After image features are detected and described as keypoint nodes of a base graph, the GNN module, which primarily aims at increasing the h-GNN’s depth, creates successive hierarchies of compressed-size graphs from the base graph through a clustering technique termed SC+PCA. SC+PCA combines Principal Component Analysis (PCA) with Spectral Clustering (SC) to enrich nodes with local and global information during graph clustering. A dual non-contrastive clustering loss is used to optimize graph clustering. Additionally, four message-passing mechanisms have been proposed to only update node representations within a graph cluster at the same hierarchical level or to update node representations across graph clusters at different hierarchical levels. The matching module performs iterative pairwise matching on the enriched node representations to obtain a scoring matrix. This matrix comprises scores indicating potential correct matches between the image keypoint nodes. The score matrix is refined with a ‘dustbin’ to further suppress unmatched features. There is a reprojection loss used to optimize keypoint match positions. The Sinkhorn algorithm generates a final partial assignment from the refined score matrix. Experimental results demonstrate the performance of the proposed h-GNN against competing state-of-the-art (SOTA) GNN-based methods on several image matching tasks under homography, estimation, indoor and outdoor camera pose estimation, and 3D reconstruction on multiple datasets. Experiments also demonstrate improved computational memory and runtime, approximately 38.1% and 26.14% lower than SuperGlue, and an average of about 6.8% and 7.1% lower than LightGlue. Future research will explore the effects of integrating more recent simplicial message-passing mechanisms, which concurrently update both node and edge representations, into our proposed model.

Introduction: Pneumonia is a microorganism infection that causes chronic inflammation of the human lung cells. Chest X-ray imaging is the most well-known screening approach used for detecting pneumonia in the early stages. While chest-Xray images are mostly blurry with low illumination, a strong feature extraction approach is required for promising identification performance. Objectives: A new hybrid explainable deep learning framework is proposed for accurate pneumonia disease identification using chest X-ray images. Methods: The proposed hybrid workflow is developed by fusing the capabilities of both ensemble convolutional networks and the Transformer Encoder mechanism. The ensemble learning backbone is used to extract strong features from the raw input X-ray images in two different scenarios: ensemble A (i.e., DenseNet201, VGG16, and GoogleNet) and ensemble B (i.e., DenseNet201, InceptionResNetV2, and Xception). Whereas, the Transformer Encoder is built based on the self-attention mechanism with multilayer perceptron (MLP) for accurate disease identification. The visual explainable saliency maps are derived to emphasize the crucial predicted regions on the input X-ray images. The end-to-end training process of the proposed deep learning models over all scenarios is performed for binary and multi-class classification scenarios. Results: The proposed hybrid deep learning model recorded 99.21% classification performance in terms of overall accuracy and F1-score for the binary classification task, while it achieved 98.19% accuracy and 97.29% F1-score for multi-classification task. For the ensemble binary identification scenario, ensemble A recorded 97.22% accuracy and 97.14% F1-score, while ensemble B achieved 96.44% for both accuracy and F1-score. For the ensemble multiclass identification scenario, ensemble A recorded 97.2% accuracy and 95.8% F1-score, while ensemble B recorded 96.4% accuracy and 94.9% F1-score. Conclusion: The proposed hybrid deep learning framework could provide promising and encouraging explainable identification performance comparing with the individual, ensemble models, or even the latest AI models in the literature. The code is available here: https://github.com/chiagoziemchima/Pneumonia_Identificaton.

Glaucoma, a predominant cause of visual impairment on a global scale, poses notable challenges in diagnosis owing to its initially asymptomatic presentation. Early identification is vital to prevent irreversible vision impairment. Cutting-edge deep learning techniques, such as vision transformers (ViTs), have been employed to tackle the challenge of early glaucoma detection. Nevertheless, limited approaches have been suggested to improve glaucoma classification due to issues like inadequate training data, variations in feature distribution, and the overall quality of samples. Furthermore, fundus images display significant similarities and slight discrepancies in lesion sizes, complicating glaucoma classification when utilizing ViTs. To address these obstacles, we introduce the contour-guided and augmented vision transformer (CA-ViT) for enhanced glaucoma classification using fundus images. We employ a Conditional Variational Generative Adversarial Network (CVGAN) to enhance and diversify the training dataset by incorporating conditional sample generation and reconstruction. Subsequently, a contour-guided approach is integrated to offer crucial insights into the disease, particularly concerning the optic disc and optic cup regions. Both the original images and extracted contours are given to the ViT backbone; then, feature alignment is performed with a weighted cross-entropy loss. Finally, in the inference phase, the ViT backbone, trained on the original fundus images and augmented data, is used for multi-class glaucoma categorization. By utilizing the Standardized Multi-Channel Dataset for Glaucoma (SMDG), which encompasses various datasets (e.g., EYEPACS, DRISHTI-GS, RIM-ONE, REFUGE), we conducted thorough testing. The results indicate that the proposed CA-ViT model significantly outperforms current methods, achieving a precision of 93.0%, a recall of 93.08%, an F1 score of 92.9%, and an accuracy of 93.0%. Therefore, the integration of augmentation with the CVGAN and contour guidance can effectively enhance glaucoma classification tasks.

The unparalleled availability of Satellite Image Time Series (SITS) for crop phenology classification unravels agricultural parcel observation and monitoring with applications of both economic and ecological importance. Moreover, the need for distinct classification of agricultural parcels into individual crop types falls on state-of-the-art deep learning models for this extrinsic task. However, most existing approaches implemented are complex and ineffective attention incorporated models, which in turn lack the resilience to recognize useful bands in achieving greater accuracy. We propose a Multi-Fast Channel Attention module for deep CNNs based on a Spatial Encoder (SE-MFCA) that requires a few parameters while enhancing the performance-complexity trade-off dilemma. Hence, we leverage on spatial encoder module to extract the images as disorderly sets of pixels to enhance the coarse spatial resolution features. We empirically show that appropriate parameter sharing in the cross channel interaction can preserve performance while significantly reducing model complexity. The proposed multi-channel attention module can efficiently be implemented via an encoder-decoder network to prevent the loss of detailed spatial information. Again, we parallelly distributed the input channel into multiple heads in our network to recover the specialized input features, which will concatenate with the residual to form a rich single feature representation. The extensive experiment has shown that our model SE-MFCA is efficient and effective compared with the previous state-of-the-art time series classification algorithm on a publicly available dataset of Sentinel-2 images for agricultural parcels. Performance-wise SE-MFCA achieves the highest overall accuracy of 94.50% and the highest mean intersection over union score of 51.92%, besides the least trainable params of 131 K and fewer floating point operations of 0.16 M.

An EEG signal (Electroencephalogram) is a bioelectric phenomenon reflecting human brain activities. In this paper, we propose a novel deep learning framework ESML (EEG-based Subject Matching Learning) using raw EEG signals to learn latent representations for EEG-based user identification and tack classification. ESML consists of two parts: one is the ESML1 model via an LSTM-based method for EEG-user linking, and one is the ESML2 model via a CNN-based method for EEG-task linking. The new model ESML is simple, but effective and efficient. It does not require any restrictions for EEG data collection on motions and thinking for users, and it does not need any EEG preprocessing operations, such as EEG denoising and feature extraction. The experiments were conducted on three public datasets and the results show that ESML performs the best and achieves significant performance improvement when compared to baseline methods (i.e., SVM, LDA, NN, DTS, Bayesian, AdaBoost and MLP). The ESML1 model provided the best precision at 96% with 109 users and the ESML2 model achieved 99% precision at 3-Class task classification. These experimental results provide direct evidence that EEG signals can be used for user identification and task classification.

Distributed learning over data from sensor-based networks has been adopted to collaboratively train models on these sensitive data without privacy leakages. We present a distributed learning framework that involves the integration of secure multi-party computation and differential privacy. In our differential privacy method, we explore the potential of output perturbation and gradient perturbation and also progress with the cutting-edge methods of both techniques in the distributed learning domain. In our proposed multi-scheme output perturbation algorithm (MS-OP), data owners combine their local classifiers within a secure multi-party computation and later inject an appreciable amount of statistical noise into the model before they are revealed. In our Adaptive Iterative gradient perturbation (MS-GP) method, data providers collaboratively train a global model. During each iteration, the data owners aggregate their locally trained models within the secure multi-party domain. Since the conversion of differentially private algorithms are often naive, we improve on the method by a meticulous calibration of the privacy budget for each iteration. As the parameters of the model approach the optimal values, gradients are decreased and therefore require accurate measurement. We, therefore, add a fundamental line-search capability to enable our MS-GP algorithm to decide exactly when a more accurate measurement of the gradient is indispensable. Validation of our models on three (3) real-world datasets shows that our algorithm possesses a sustainable competitive advantage over the existing cutting-edge privacy-preserving requirements in the distributed setting.

Multiple organizations would benefit from collaborative learning models trained over aggregated datasets from various human activity recognition applications without privacy leakages. Two of the prevailing privacy-preserving protocols, secure multi-party computation and differential privacy, however, are still confronted with serious privacy leakages: lack of provision for privacy guarantee about individual data and insufficient protection against inference attacks on the resultant models. To mitigate the aforementioned shortfalls, we propose privacy-preserving architecture to explore the potential of secure multi-party computation and differential privacy. We utilize the inherent prospects of output perturbation and gradient perturbation in our differential privacy method, and progress with an innovation for both techniques in the distributed learning domain. Data owners collaboratively aggregate the locally trained models inside a secure multi-party computation domain in the output perturbation algorithm, and later inject appreciable statistical noise before exposing the classifier. We inject noise during every iterative update to collaboratively train a global model in our gradient perturbation algorithm. The utility guarantee of our gradient perturbation method is determined by an expected curvature relative to the minimum curvature. With the application of expected curvature, we theoretically justify the advantage of gradient perturbation in our proposed algorithm, therefore closing existing gap between practice and theory. Validation of our algorithm on real-world human recognition activity datasets establishes that our protocol incurs minimal computational overhead, provides substantial utility gains for typical security and privacy guarantees.

ABSTRACT: Detecting and classifying animal species is the first step in determining their long-term viability and the influence we may be having on them. Second, it aids people in recognizing predators and non-predatory animals, both of which pose a significant threat to humans and the environment. Third, it lowers the rate of traffic accidents in various regions since it has been a regular sighting on roadways, resulting in several collisions with automobiles. However, animal species' detection and Classification of animal species face many challenges such as the size and inconsistent behaviors various among the species. This paper proposes using a novel two-stage network with a modified multi-scale attention mechanism to create a more integrated recognition and classification system to attend to the challenges. At the regional proposal stage, a deeply characterized pyramid design with lateral connections was adopted, making the semantic characteristic of a small item more sensitive. Secondly, by reason of a densely connected convolutional network, the functional transmission is enhanced and multiplexed throughout the classification stage, resulting in a more precise Classification with fewer parameters. The Proposed model was evaluated using the AP and mAP evaluation metrics on the Animal wildlife and the challenging Animal-80 dataset. An mAP of +0.1% and an AP of 5% to 20% increase in each class was achieved by the attention-based proposed model compared to the non-attention-based model. Further comparison with other related works shows the proposed techniques' effectiveness for detecting and classifying animal species.

Intelligent transportation systems (ITSs) have become an indispensable component of modern global technological development, as they play a massive role in the accurate statistical estimation of vehicles or individuals commuting to a particular transportation facility at a given time. This provides the perfect backdrop for designing and engineering an adequate infrastructural capacity for transportation analyses. However, traffic prediction remains a daunting task due to the non-Euclidean and complex distribution of road networks and the topological constraints of urbanized road networks. To solve this challenge, this paper presents a traffic forecasting model which combines a graph convolutional network, a gated recurrent unit, and a multi-head attention mechanism to simultaneously capture and incorporate the spatio-temporal dependence and dynamic variation in the topological sequence of traffic data effectively. By achieving 91.8% accuracy on the Los Angeles highway traffic (Los-loop) test data for 15-min traffic prediction and an R2 score of 85% on the Shenzhen City (SZ-taxi) test dataset for 15- and 30-min predictions, the proposed model demonstrated that it can learn the global spatial variation and the dynamic temporal sequence of traffic data over time. This has resulted in state-of-the-art traffic forecasting for the SZ-taxi and Los-loop datasets.

Over the years, Steganography and Cryptography have been complementary techniques for enforcing security of digital data. The need for the development of robust multi-layered schemes to counter the exponential grow in the power of computing devices that can compromise security is critical in the design and implementation of security systems. Therefore, we propose a new combined steganographic and cryptographic scheme using the operators of genetic algorithm (GA) such selection, crossover and mutation, and some properties of the residue number system (RNS) with an appropriate fusing technique in order to embed encrypted text within images. The proposed scheme was tested using MatLab® R2017b and a CORE™i7 processor. Simulation results show that the proposed scheme can be deployed at one level with only the stego image containing the encrypted hidden message and at another level where the stego message is further encrypted. An analysis based on standard key metrics such as visual perception and statistical methods on steganalysis and cryptanalysis show that the proposed scheme is robust, is not complex with reduced runtime and will consume less power due to the use of residue numbers when compared to similar existing schemes.

Classical Artificial Neural Networks (ANNs) though well exploited in solving classification problems, do not model perfectly the information encoding process in the human brain because ANNs encode information using rate-based coding. However, biological neurons in the brain are known to encode information using temporal coding. In order to mimic the biological method of encoding information, various Spiking Neural Network (SNN) models have been developed. However, some of these models are limited in the number of spikes and do not leverage well on some classification problems. In order to address some of the inherent challenges associated with SNN, a multi-layer learning model for a multi-spiking network is proposed in this paper. The model exploits the temporal coding of spikes and the least-squares method to derive a weight update scheme. It also employs a spike locality concept in order to determine how the synaptic weights are to be adjusted at a particular spike time so as to minimize the learning interference, and thereby, increasing the number of spikes for learning. The performance of the model is evaluated on benchmarked classification datasets. A correlation-based metric is combined with a threshold concept to measure the classification accuracy of the model. The experimental results showed that the proposed model achieved better classification accuracy than some state-of-the-art multi-layer SNN learning models.

Secure localization of vehicles is gaining the attention of researchers from both academia and industry especially due to the emergence of internet of things (IoT). The modern vehicles are usually equipped with circuitries that gives connectivity with other vehicles and with cellular networks such as 4G/Fifth generation cellar network (5G). The challenge of secure localization and positioning is magnified further with the invention of technologies such as autonomous or driverless vehicles based on IoT, satellite, and 5G. Some satellite and IoT based localization techniques exploit machine learning, semantic segmentation, and access control mechanism. Access control provides access grant and secure information sharing mechanism to authorized users and restricts unauthorized users, which is necessary regarding security and privacy of government or military vehicles. Previously, static conflict of interest (COI) based access control was used for security proposes. However, static COI based access control creates excesses and administrative overload that creates latency in execution, which is the least tolerable factor in modern IoT or 5G control vehicles. Therefore, in this paper, a hybrid access control (HAC) model is proposed that implements the dynamic COI in the HAC model on the level of roles. The proposed model is enhanced by modifying the role-based access control (RBAC) model by inserting new attributes of the RBAC entities. The HAC model deals with COI on the level of roles in an efficient manner as compared to previously proposed models. Moreover, this model features significant improvement in terms of dynamic behavior, decreased administrative load, and security especially for vehicular localization. Furthermore, the mathematical modeling of the proposed model is implemented with an example scenario to validate the concept.

Due to the complexity of the anatomical structure for human organs, medical image segmentation is always a challenging computer vision task. The Convolutional Neural Network (CNN) requires a rich feature representation, which not only needs the convolutional layers from shallow to deep,but also requires the resolution from small to large. Although CNN can be used to fuse mid-level features that are employed short-cutting, this just is a simple “shallow” connection. Thus, how to obtain useful features and how to utilize these features to improve the segmentation processes are still the key issues. In this paper, Multi-features Refinement and Aggregation (MRA) makes full use of hierarchical features by using the features fusion on several levels, and reveal the importance of refinement and aggregation of features in the medical image segmentation process. The network get low-level, high-level and even mid-level features by sampling. After aggregation and re-extraction, these features are more effectively combined. Experiment results show that our method can significantly improve segmentation accuracy compared to existing feature fusion schemes. And our approach is generalized to different backbone networks with consistent accuracy gain in brain segmentation, and it sets a new state-of-the-art in the Brat-2015 benchmarks.

Deep Neural Networks (DNNs) classifiers performance degrades under adversarial attacks, such attacks are indistinguishably perturbed relative to the original data. Providing robustness to adversarial attacks is an important challenge in DNN training, which has led to extensive research. In this paper, we harden DNN classifiers under the adversarial attacks by regularizing their deep internal representation space with Multi-class Triplet regularization method. This method enables DNN classifier to learn a feature representation that detects similarities between adversarial and clean images and brings similar images close to their original class and pushes dissimilar images away from their false classes. This training process with our Multi-class Triplet regularization method in combination with Gaussian noise injection proves to be more robust in detecting adversarial attacks exceeding that of adversarial training on strong iterative attacks.

Public Key Encryption with Keyword Search (PEKS) is a desirable technique to provide searchable functionality over encrypted data in public key settings, which allows a user to delegate a third party server to perform the search operation on encrypted data by means of keyword search trapdoor without learning about the data. However, the existing PEKS schemes cannot be directly applied to practice due to keyword guessing attack or the absence of a mechanism to limit the lifetime of a trapdoor. By addressing these issues at the same time, this paper presents a Public Key Encryption Scheme with Temporary and Fuzzy Keyword Search (PETFKS) by using a fuzzy function and an encryption tree. The proposed PETFKS scheme is proven adaptively secure concerning keyword confidentiality and backward and forward secrecy in the random oracle model under the Bilinear Di e-Hellman assumption. Moreover, it is also proven selectively secure with regard to the resistance of keyword guessing attack. Furthermore, the security and e ciency analyses of the proposed scheme are provided by comparing to the related works. The analyses indicate that the proposed scheme makes a threefold contribution to the practical application of public key encryption with keyword search, namely o ering secure search operation, limiting the lifetime of a trapdoor and enabling secure time-dependent data retrieval.

According to research, classifiers and detectors are less accurate when images are blurry, have low contrast, or have other flaws which raise questions about the machine learning model’s ability to recognize items effectively. The chest X-ray image has proven to be the preferred image modality for medical imaging as it contains more information about a patient. Its interpretation is quite difficult, nevertheless. The goal of this research is to construct a reliable deep-learning model capable of producing high classification accuracy on chest x-ray images for lung diseases. To enable a thorough study of the chest X-ray image, the suggested framework first derived richer features using an ensemble technique, then a global second-order pooling is applied to further derive higher global features of the images. Furthermore, the images are then separated into patches and position embedding before analyzing the patches individually via a vision transformer approach. The proposed model yielded 96.01% sensitivity, 96.20% precision, and 98.00% accuracy for the COVID-19 Radiography Dataset while achieving 97.84% accuracy, 96.76% sensitivity and 96.80% precision, for the Covid-ChestX-ray-15k dataset. The experimental findings reveal that the presented models outperform traditional deep learning models and other state-of-the-art approaches provided in the literature.

Today, the trend of the Internet of Things (IoT) is increasing through the use of smart devices, vehicular networks, and household devices with internet-based networks. Specifically, the IoT smart devices and gadgets used in government and military are crucial to operational success. Communication and data sharing between these devices have increased in several ways. Similarly, the threats of information breaches between communication channels have also surged significantly, making data security a challenging task. In this context, access control is an approach that can secure data by restricting unauthorized users. Various access control models exist that can effectively implement access control yet, and there is no single state-of-the-art model that can provide dynamicity, security, ease of administration, and rapid execution all at once. In combating this loophole, we propose a novel secure and dynamic access control (SDAC) model for the IoT networks (smart traffic control and roadside parking management). Our proposed model allows IoT devices to communicate and share information through a secure means by using wired and wireless networks (Cellular Networks or Wi-Fi). The effectiveness and efficiency of the proposed model are demonstrated using mathematical models and discussed with many example implementations.

The ongoing coronavirus 2019 (COVID-19) pandemic caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has resulted in a severe ramification on the global healthcare system, principally because of its easy transmission and the extended period of the virus survival on contaminated surfaces. With the advances in computer-aided diagnosis and artificial intelligence, this paper presents the application of deep learning and adversarial network for the automatic identification of COVID-19 pneumonia in computed tomography (CT) scans of the lungs. The complexity and time limitation of the reverse transcription-polymerase chain reaction (RT-PCR) swab test makes it disadvantageous to depend solely on as COVID-19’s central diagnostic mechanism. Since CT imaging systems are of low cost and widely available, we demonstrate that the drawback of the RT-PCR can be alleviated with a faster, automated, and reduced contact diagnostic process via the use of a neural network model for the classification of infected and noninfected CT scans. In our proposed model, we explore the benefit of transfer learning as a means of resolving the problem of inadequate dataset and the importance of semisupervised generative adversarial network for the extraction of well-mapped features and generation of image data. Our experimental evaluation indicates that the proposed semisupervised model achieves reliable classification, taking advantage of the reflective loss distance between the real data sample space and the generated data.

Vehicular ad-hoc networks (VANETs) are the crucial part of intelligent transportation systems (ITS), which are brought to enhance the security, efficiency, and comfort of transportation. VANETs have aroused extensive attention in the world recently. One of the challenges in practice is real time and low delay, which strongly requires VANETs to be efficient. Existing schemes have properly solved the problem which is how to aggregate the signatures and verify the aggregated signature. However, few solutions are proposed to pinpoint all invalid signatures if existing. The algorithms that can find all invalid signatures are not efficient enough. Following consideration of the above deficiencies of existing approaches, this paper proposes a factorial bitwise divisions (FBD) algorithm and its optimized version and early-stopping factorial bitwise divisions (EFBD) algorithm. Both algorithms are parallel-friendly. Compared with the binary-based batch verification algorithm, the experimental results demonstrate that the proposed algorithms achieve better performance in both theory and practice at low invalid signatures’ rate. Especially, in the parallel condition, when the number of invalid signatures is 1, the proposed algorithms cost only one aggregation-verification delay, while the comparison is more than log2 n times.

With the emergence of deep learning, computer vision has witnessed extensive advancement and has seen immense applications in multiple domains. Specifically, image captioning has become an attractive focal direction for most machine learning experts, which includes the prerequisite of object identification, location, and semantic understanding. In this paper, semantic segmentation and image captioning are comprehensively investigated based on traditional and state-of-the-art methodologies. In this survey, we deliberate on the use of deep learning techniques on the segmentation analysis of both 2D and 3D images using a fully convolutional network and other high-level hierarchical feature extraction methods. First, each domain’s preliminaries and concept are described, and then semantic segmentation is discussed alongside its relevant features, available datasets, and evaluation criteria. Also, the semantic information capturing of objects and their attributes is presented in relation to their annotation generation. Finally, analysis of the existing methods, their contributions, and relevance are highlighted, informing the importance of these methods and illuminating a possible research continuation for the application of semantic image segmentation and image captioning approaches.

Brain tumor segmentation plays an important role in diagnosing brain tumor. Nowadays, intense interest has been received in applying convolution neural networks in medical image analysis, but its performance is restricted by the limitation of the depth of the network. And how to accelerate the information propagation and make full use of all the hierarchical features in the network is also of vital importance. To address these problems, this paper proposed Deep Residual Dilate Network with Middle Supervision (RDM-Net), which combines the residual network with dilated convolution. It can solve the problem of vanishing gradient and increase the receptive field without reducing the resolution. During images processing, some information of regions of small tumor could be discarded, for its resolution is attenuated to a single pixel by continuously convolutional operations. Therefore, the spatial fusion block, consisting of a pixel discriminator and a region discriminator, has been designed to reserve the detailed information in the region of small tumor. It evaluates the relationship between this single pixel and its adjacent region to obtain the spatial structure information of brain tumors. Furthermore, the middle supervision block consisting of proposal pyramid and multi-hierarchical loss is proposed, which shortens the distance of information path and reduces cumulative errors during the network training. The proposal pyramid is inspired by the idea of boost learning, which fuses each proposal at multiple resolution level to ensure that the network produces better predictions. Multi-hierarchical loss combines the loss of intermediate proposal in the middle layers and the loss of prediction of the output layer to achieve the effect of middle supervision. The results of experiments illustrate that our framework can effectively propagate features of each layer and increase the diversity of information to enhance feature hierarchy for medical image recognition. Compared to other state-of-the-art methods, our framework has performed well in the BRATS2015 challenge. In summary, the main contribution in our paper is that this work is an early attempt to adopt the concept of “middle supervision” on multi-modal brain tumor segmentation.

User-provided tags associated with social images are essential information for social image retrieval. Unfortunately, these tags are often imperfect to describe the visual contents, which severely degrades the performance of image retrieval. Tag relevance learning models are proposed to improve the descriptive powers of tags mostly based on the Gaussian noise assumption. However, the intrinsic probability distribution of the noise is unknown and other probability distributions may be much better. Towards this end, this paper investigates the applicable probability distributions of tag noise and proposes a novel Cauchy Matrix Factorization (CMF) method for tag-based image retrieval. The Cauchy probability distribution is robust to all kinds of noise and more suitable to model the tagging noise of social images. Therefore, we utilize Cauchy distribution to model noise under the matrix factorization framework. Besides, other five probability density functions, i.e., Gaussian, Laplacian, Poisson, Student-t and Logistic, are investigated to model noise of social tags. To evaluate the performance of different probability distributions, extensive experiments on two widely-used datasets are conducted and results show the robustness of CMF to noisy tags of social images.

As a novel network architecture, information-centric networking (ICN) has a good performance in facilitating content sharing among users. Although in-network cache used in ICN can effectively solve the problem of network congestion. Meanwhile, it also brings a lot of challenges such as security and privacy issues during data transformation. Some of the current solutions are based on the traditional encryption technology, but these solutions introduce significant overhead in the client side and have a high requirement to the memory and computing power of end user. In this paper, we use an efficient proxy re-encryption (PRE) scheme in ICN framework to help reduce the overhead on the user-side while guaranteeing flexible data sharing between subscribers and even their cooperator. Our proposal has the additional benefits of a non-interactivity and collusion resistance. We also prove that our scheme is secure against adaptive replayable adaptive chosen ciphertext attack (RCCA) in re-encryption and chosen ciphertext attack (CCA) secure in complete ICN encryption. Our analysis of this program also shows that the scheme has a relatively good performance in computation cost and communication complexity aspects.

The generation of the textual description of the differences in images is a relatively new concept that requires the fusion of both computer vision and natural language techniques. In this paper, we present a novel Fully Convolutional CaptionNet (FCC) that employs an encoder-decoder framework to perform visual feature extractions, compute the feature distances, and generate new sentences describing the measured distances. After extracting the features of the images, a contrastive function is used to compute their weighted L1 distance which is learned and selectively attended to determine salient sections of the feature at every time step. The attended feature region is adequately matched to corresponding words iteratively until a sentence is completed. We propose the application of upsampling network to enlarge the features' field of view, this provides a robust pixel-based discrepancy computation. Our extensive experiments indicate that the FCC model outperforms other learning models on the benchmark Spot-the-Diff datasets by generating succinct and meaningful textual differences in images.

Privacy in the Internet of Things is a fundamental challenge for the Ubiquitous healthcare systems that depend on the data aggregated and collaborative deep learning among different parties. This paper proposes the MSCryptoNet, a novel framework that enables the scalable execution and the conversion of the state-of-the-art learned neural network to the MSCryptoNet models in the privacy-preservation setting. We also design a method for approximation of the activation function basically used in the convolutional neural network (i.e., Sigmoid and Rectified linear unit) with low degree polynomials, which is vital for computations in the homomorphic encryption schemes. Our model seems to target the following scenarios: 1) the practical way to enforce the evaluation of classifier whose inputs are encrypted with possibly different encryption schemes or even different keys while securing all operations including intermediate results and 2) the minimization of the communication and computational cost of the data providers. The MSCryptoNet is based on the multi-scheme fully homomorphic encryption. We also prove that the MSCryptoNet as a privacy-preserving deep learning scheme over the aggregated encrypted data is secured.

Device-to-Device (D2D) communication is a promising method for the emerging Internet of Things. Secure information exchange plays a key role in the application of D2D communication. Considering that the wireless devices are powered by batteries, in this paper, a lightweight secure D2D system is designed by using multiple sensors on mobile devices. Specifically, by leveraging an acceleration sensor equipped in two wireless devices, a lightweight and efficient key distribution scheme for secure D2D communication is proposed. Based on the distributed secure key, an efficient near-field authentication is developed with a speaker and a microphone to determine whether these two devices are physically close; and a secure information exchange scheme with high efficiency, which includes message encryption/decryption and message authentication, is presented over the audio channel and the RF channel. The Extensive experiments are provided to demonstrate that our system can achieve a secure information exchange between two wireless devices with low energy consumption and computing resources.

Introduction and Background: Despite fast developments in the medical field, histological diagnosis is still regarded as the benchmark in cancer diagnosis. However, the input image feature extraction that is used to determine the severity of cancer at various magnifications is harrowing since manual procedures are biased, time consuming, labor intensive, and error-prone. Current state-of-the-art deep learning approaches for breast histopathology image classification take features from entire images (generic features). Thus, they are likely to overlook the essential image features for the unnecessary features, resulting in an incorrect diagnosis of breast histopathology imaging and leading to mortality. Methods: This discrepancy prompted us to develop DEEP_Pachi for classifying breast histopathology images at various magnifications. The suggested DEEP_Pachi collects global and regional features that are essential for effective breast histopathology image classification. The proposed model backbone is an ensemble of DenseNet201 and VGG16 architecture. The ensemble model extracts global features (generic image information), whereas DEEP_Pachi extracts spatial information (regions of interest). Statistically, the evaluation of the proposed model was performed on publicly available dataset: BreakHis and ICIAR 2018 Challenge datasets. Results: A detailed evaluation of the proposed model’s accuracy, sensitivity, precision, specificity, and f1-score metrics revealed the usefulness of the backbone model and the DEEP_Pachi model for image classifying. The suggested technique outperformed state-of-the-art classifiers, achieving an accuracy of 1.0 for the benign class and 0.99 for the malignant class in all magnifications of BreakHis datasets and an accuracy of 1.0 on the ICIAR 2018 Challenge dataset. Conclusions: The acquired findings were significantly resilient and proved helpful for the suggested system to assist experts at big medical institutions, resulting in early breast cancer diagnosis and a reduction in the death rate.

Today, accurate and automated abnormality diagnosis and identification have become of paramount importance as they are involved in many critical and life-saving scenarios. To accomplish such frontiers, we propose three artificial intelligence models through the application of deep learning algorithms to analyze and detect anomalies in human heartbeat signals. The three proposed models include an attention autoencoder that maps input data to a lower-dimensional latent representation with maximum feature retention, and a reconstruction decoder with minimum remodeling loss. The autoencoder has an embedded attention module at the bottleneck to learn the salient activations of the encoded distribution. Additionally, a variational autoencoder (VAE) and a long short-term memory (LSTM) network is designed to learn the Gaussian distribution of the generative reconstruction and time-series sequential data analysis. The three proposed models displayed outstanding ability to detect anomalies on the evaluated five thousand electrocardiogram (ECG5000) signals with 99% accuracy and 99.3% precision score in detecting healthy heartbeats from patients with severe congestive heart failure.

Personal health record (PHR) is a patient-centric model of health information exchange, which greatly facilitates the storage, access, and share of personal health information. In order to share the valuable resources and reduce the operational cost, the PHR service providers would like to store the PHR applications and health information data in the cloud. The private health information may be exposed to unauthorized organizations or individuals since the patient lost the physical control of their health information. Ciphertext-policy attribute-based signcryption is a promising solution to design a cloud-assisted PHR secure sharing system. It provides fine-grained access control, confidentiality, authenticity, and sender privacy of PHR data. However, a large number of pairing and modular exponentiation computations bring heavy computational overhead during designcryption process. In order to reconcile the conflict of high computational overhead and low efficiency in the designcryption process, an outsourcing scheme is proposed in this paper. In our scheme, the heavy computations are outsourced to ciphertext transformed server, only leaving a small computational overhead for the PHR user. At the same time, the extra communication overhead in our scheme is actually tolerable. Furthermore, theoretical analysis and the desired securing properties including confidentiality, unforgeability, and verifiability have been proved formally in the random oracle model. Experimental evaluation indicates that the proposed scheme is practical and feasible.

Public key encryption supporting equality test (referred to as PKE-ET) provides the capability of testing the equivalence between two messages encrypted under different public keys. Ciphertext-policy attribute-based encryption (CP-ABE) is a promising primitive to achieve versatile and secure data sharing in the cloud computing by providing flexible one-to-many encryption. In this paper, we first initialize the concept of CP-ABE with equality test (CP-ABE-ET) by combining the notions of PKE-ET and CP-ABE. Using ABE-ET primitive, the receiver can delegate a cloud server to perform an equivalence test between two messages, which are encrypted under different access policies. During the delegated equivalence test, the cloud server is unable to obtain any knowledge of the message encrypted under either access policy. We propose a concrete CP-ABE-ET scheme using bilinear pairing and Vi`ete's formulas, and give the security proof of the proposed scheme formally in the standard model. Moreover, the theoretic analysis and experimental simulation reveal that the proposed scheme is efficient and practical.

Fog computing dramatically extends the cloud computing to the edge of the network and admirably solves the problem that the brokers (in publish-subscribe system) generally lack of computing capacity and energy power. However, brokers may be disguised, hacked, sniffed, and corrupted. The traditional security technology cannot protect the system privacy when facing a possible collusion attack. In this paper, we propose a privacy-preserving content-based publish/subscribe scheme with differential privacy in fog computing context, named PCP, where the fog nodes act as the brokers. Specifically, PCP firstly utilizes the U-Apriori algorithm to mine the top-K frequent itemsets (i.e., the attributes) from uncertain data sets, then applies the exponential and Laplace mechanism to ensure the differential privacy, and the broker uses the mined top-K itemsets to match appropriate publisher and subscriber finally. Security analysis shows that the PCP can guarantee differential privacy in theory. To evaluate the performance of PCP, we carry out experiments with real-world scenario data sets. The experimental results show that PCP efficiently achieves the tradeoff between the system cost and the privacy demand.

Due to massive mobile terminal devices and ubiquitous communication, the Internet of things (IoT) has become an inevitable trend. Given that the fifth generation (5G) wireless networks expects to drive the proliferation of the IoT and may extend the access functions and systems of the IoT, it makes the IoT a vitally important part in future 5G wireless networks. Simultaneously, the limit of the bandwidth and power of the 5G would adversely affect the widespread promotion of the IoT. However, wireless caching techniques could remarkably resolve this issue. Recently, using fog nodes to improve the capacity of caching has become a trend in caching system. However, node-based caching systems may suffer from malicious access and destruction. To protect caching from sabotage and to further ensure its reliability, we propose a new lightweight label-based access control scheme (LACS) that authenticates the authorized fog nodes to ensure protection. Specifically, the LACS can authenticate the fog nodes by verifying the integrity of the shared files that are embedded label values, and only the authenticated fog nodes can access the caching service. The analysis shows that the proposed scheme is verifiable (the malicious fog node cannot cheat the caching server to pretend to be a legal node) and efficient in both computation and verification. Moreover, simulation experiments show that the LACS can reach the millisecond-level verification and it has a good accuracy.

Forecasting the direction and trend of stock price is an important task which helps investors to make prudent financial decisions in the stock market. Investment in the stock market has a big risk associated with it. Minimizing prediction error reduces the investment risk. Machine learning (ML) models typically perform better than statistical and econometric models. Also, ensemble ML models have been shown in the literature to be able to produce superior performance than single ML models. In this work, we compare the effectiveness of tree-based ensemble ML models (Random Forest (RF), XGBoost Classifier (XG), Bagging Classifier (BC), AdaBoost Classifier (Ada), Extra Trees Classifier (ET), and Voting Classifier (VC)) in forecasting the direction of stock price movement. Eight different stock data from three stock exchanges (NYSE, NASDAQ, and NSE) are randomly collected and used for the study. Each data set is split into training and test set. Ten-fold cross validation accuracy is used to evaluate the ML models on the training set. In addition, the ML models are evaluated on the test set using accuracy, precision, recall, F1-score, specificity, and area under receiver operating characteristics curve (AUC-ROC). Kendall W test of concordance is used to rank the performance of the tree-based ML algorithms. For the training set, the AdaBoost model performed better than the rest of the models. For the test set, accuracy, precision, F1-score, and AUC metrics generated results significant to rank the models, and the Extra Trees classifier outperformed the other models in all the rankings.

Role-based access control (RBAC) continues to gain popularity in the management of authorization concerning access to knowledge assets in organizations. As a socio-technical concept, the notion of role in RBAC has been overemphasized, while very little attention is given to the precursors: role strain, role ambiguity, and role conflict. These constructs provide more significant insights into RBAC design in Knowledge Management Systems (KMS). KMS is the technology-based knowledge management tool used to acquire, store, share, and apply knowledge for improved collaboration and knowledge-value creation. In this paper, we propose eight propositions that require future research concerning the RBAC system for knowledge security. In addition, we propose a model that integrates these precursors and RBAC to deepen the understanding of these constructs. Further, we examine these precursory constructs in a socio-technical fashion relative to RBAC in the organizational context and the status–role relationship effects. We carried out conceptual analysis and synthesis of the relevant literature, and present a model that involves the three essential precursors that play crucial roles in role mining and engineering in RBAC design. Using an illustrative case study of two companies where 63 IT professionals participated in the study, the study established that the precursors positively and significantly increase the intractability of the RBAC system design. Our framework draws attention to both the management of organizations and RBAC system developers about the need to consider and analyze the precursors thoroughly before initiating the processes of policy engineering, role mining, and role engineering. The propositions stated in this study are important considerations for future work.

Access control has become problematic in several organizations because of the difficulty in establishing security and preventing malicious users from mimicking roles. Moreover, there is no flexibility among users in the participation in their roles, and even controlling them. Several role-based access control (RBAC) mechanisms have been proposed to alleviate these problems, but the security has not been fully realized. In this work, however, we present an RBAC model based on blockchain technology to enhance user authentication before knowledge is accessed and utilized in a knowledge management system (KMS). Our blockchain-based system model and the smart contract ensure that transparency and knowledge resource immutability are achieved. We also present smart contract algorithms and discussions about the model. As an essential part of RBAC model applied to KMS environment, trust is ensured in the network. Evaluation results show that our system is efficient.

Aspect-level sentiment analysis has drawn growing attention in recent years, with higher performance achieved through the attention mechanism. Despite this, previous research does not consider some human psychological evidence relating to language interpretation. This results in attention being paid to less significant words especially when the aspect word is far from the relevant context word or when an important context word is found at the end of a long sentence. We design a novel model using word significance to direct attention towards the most significant words, with novelty decay and incremental interpretation factors working together as an alternative for position based models. The interpretation factor represents the maximization of the degree each new encountered word contributes to the sentiment polarity and a counter balancing stretched exponential novelty decay factor represents decaying human reaction as a sentence gets longer. Our findings support the hypothesis that the attention mechanism needs to be applied to the most significant words for sentiment interpretation and that novelty decay is applicable in aspect-level sentiment analysis with a decay factor β = 0.7 .

Alzheimer's disease (AD) is a neurological condition that is recognized as one of the primary causes of memory loss. AD currently has no cure. Therefore, the need to develop an efficient model with high precision for timely detection of the disease is very essential. When AD is detected early, treatment would be most likely successful. The most often utilized indicators for AD identification are the Mini-mental state examination (MMSE), and the clinical dementia. However, the use of these indicators as ground truth marking could be imprecise for AD detection. Researchers have proposed several computer-aided frameworks and lately, the supervised model is mostly used. In this study, we propose a novel 3D Convolutional Neural Network Multilayer Perceptron (3D CNN-MLP) based model for AD classification. The model uses Attention Mechanism to automatically extract relevant features from Magnetic Resonance Images (MRI) to generate probability maps which serves as input for the MLP classifier. Three MRI scan categories were considered, thus AD dementia patients, Mild Cognitive Impairment patients (MCI), and Normal Control (NC) or healthy patients. The performance of the model is assessed by comparing basic CNN, VGG16, DenseNet models, and other state of the art works. The models were adjusted to fit the 3D images before the comparison was done. Our model exhibited excellent classification performance, with an accuracy of 91.27% for AD and NC, 80.85% for MCI and NC, and 87.34% for AD and MCI. [ABSTRACT FROM AUTHOR]