Bohan Lin, Kuo Yang, Zelin Tan, Yingchuan Lai, Chen Zhang, Guibin Zhang, Xinlei Yu, Miao Yu, Xu Wang, Yudong Zhang, Yang Wang. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

Accurate cellular traffic prediction is crucial for the rational allocation of network resources. Existing methods generally overlook the accurate extraction and full utilization of features across different periods of cellular traffic. Thus, we propose a Relative Long Short-Term Adaptive Spatial-Temporal Targeted Extraction (RLSASTTE) network, which fully extracts long-term trends and short-term dynamics, and then maximizes their utilization. First, to extract long-term spatial-temporal features, we exploit the strength of the Transformer in capturing long-range temporal dependencies, address its limitations in spatial relationship modeling by redesigning a spatial-temporal attention mechanism, and further adopt a pre-decomposition strategy to emphasize seasonal component mining. Second, to capture short-term spatial-temporal features, we define a multi-dilation convolution to explore short-term dependencies and design a novel local strong correlation dynamic attention mechanism to investigate local spatial influences, which also eliminates the limitation of fixed neighboring nodes. Finally, we innovatively employ an adaptive dual gating mechanism to efficiently integrate diverse features. We conducted experiments on three real-world datasets. Compared to the state-of-the-art methods, RLSASTTE achieves reductions of at least 3.1% and 3.7% in mean absolute error and root mean square error, respectively. In addition, we explore auxiliary training based on data augmentation, achieving additional performance improvements of 2.0% to 4.5%.

Deformable image registration (DIR) is critical for accurate clinical diagnosis and effective treatment planning. However, patient movement, significant intensity differences, and large breathing deformations hinder accurate anatomical alignment in multi-modal image registration. These factors exacerbate the entanglement of anatomical and modality-specific style information, thereby severely limiting the performance of multi-modal registration. To address this, we propose a novel LAtent Diffusion-based Domain-Adaptive feature disentangling (LADDA) framework for unsupervised multi-modal medical image registration, which explicitly addresses the representation disentanglement. First, LADDA extracts reliable anatomical priors from the Latent Diffusion Model (LDM), facilitating downstream content-style disentangled learning. A Domain-Adaptive Feature Disentangling (DAFD) module is proposed to promote anatomical structure alignment further. This module disentangles image features into content and style information, boosting the network to focus on cross-modal content information. Next, a Neighborhood-Preserving Hashing (NPH) is constructed to further perceive and integrate hierarchical content information through local neighbourhood encoding, thereby maintaining cross-modal structural consistency. Furthermore, a Unilateral-Query-Frozen Attention (UQFA) module is proposed to enhance the coupling between upstream prior and downstream content information. The feature interaction within intra-domain consistent structures improves the fine recovery of detailed textures. The proposed framework is extensively evaluated on large-scale multi-center datasets, demonstrating superior performance across diverse clinical scenarios and strong generalization on out-of-distribution (OOD) data.

ABSTRACT Dragon boat racing, a popular aquatic folklore team sport, is traditionally held during the Dragon Boat Festival. Inspired by this event, we propose a novel human‐based meta‐heuristic algorithm called dragon boat optimization (DBO) in this paper. It models the unique behaviours of each crew member on the dragon boat during the race by introducing social psychology mechanisms (social loafing, social incentive). Throughout this process, the focus is on the interaction and collaboration among the crew members, as well as their decision‐making in various situations. During each iteration, DBO implements different state updating strategies. By accurately modelling the crew's behaviour and employing adaptive state update strategies, DBO consistently achieves high optimization performance, as validated by comprehensive testing on 29 benchmark functions and 2 structural design problems. Experimental results indicate that DBO outperforms 7 and 16 state‐of‐the‐art meta‐heuristic algorithms across these test functions and problems, respectively.

Amid the ongoing pandemic, reducing reliance on manual diagnostic procedures has become crucial. In this light, deep neural networks (DNNs) have demonstrated substantial progress in coronavirus disease 2019 (COVID-19) detection. However, when exposed to ‘toxic samples’—imprecise or uncertain data such as outliers, noisy or mislabeled entries, that negatively impact the training process—existing methods cannot effectively protect the training convergence from the overshoot phenomenon. This situation would slow the training convergence. Additionally, current diagnostic models necessitate substantial re-tuning time to adapt to new virus strains or to handle data from different platforms. This research focuses on the parameter updates design and propose a highly controllable and stable DNN for COVID-19 detection. By exploiting the past, current and future changes of the gradient in a fuzzy logic manner, and taking into account the cross-coupling effect between the gradient and its rate of change, we achieve dynamic, high-precision control on parameter updates in DNN optimization to reach a stable status at a faster convergence rate. In each iteration, the current learning rate adjusts itself to the current optimal value within the fuzzy neighboring region. Potentially hereditary module sequentially transfers the trained knowledge between estimators while updating the fuzzy universe range based on the calculated contraction–expansion factors. Consequently, our proposed algorithm alleviates the overshoot suffered by toxic samples, meanwhile effectively enhancing the model robustness, resource-efficiency, flexibility, adaptability, and compatibility. When tested on popular DNN architectures , it yields up to 47.18% acceleration with promising accuracy on four public datasets. Extensive experiments prove the effectiveness of our method in comparison to state-of-the-art optimizers and diagnosis systems, facilitating the real-life demands for COVID-19 detection.

Electrocardiogram (ECG) arrhythmia identification is critical for early cardiovascular disease diagnosis and monitoring in Internet of Things (IoT) industry. Still, it is difficult due to complicated waveforms, individual variability, and the requirement for real-time analysis on resource-limited equipment. Traditional approaches sometimes fail to detect complicated spatial-temporal correlations in ECG data, limiting their efficiency in identifying arrhythmias. Furthermore, deploying these models in tinyML contexts, such as edge and IoT devices limited by large computational and memory needs, emphasizes the importance of lightweight, accurate models for real-time applications. Our suggested solution consists of three main components: SigNet, DualGCNN, and FusionGCNN. SigNet uses Separable Convolution layers to effectively extract local spatial features, making it ideal for IoT-based healthcare deployment. DualGCNN combines dual Graph Convolutional layers with spatial attention, allowing the model to capture local and global dependencies for better classification of arrhythmia. FusionGCNN combines the capabilities of GCN and SigNet with an effective feature fusion technique to improve feature representation while remaining computationally economical. Ablation tests show that FusionGCNN improves performance considerably, with greater accuracy (0.9641), lower training error (0.0004), and a higher F1 Score (0.9645) across a variety of ECG patterns. FusionGCNN, with its low training error, high stability, and computational economy, is well-suited to tinyML requirements, allowing implementation on edge and IoT devices for scalable, real-time ECG monitoring in healthcare.

Advancements in 5G technology are accelerating the adoption of multiaccess edge computing (MEC), and the widespread deployment of high-density edge devices has introduced significant energy consumption challenges. Edge demand response (EDR) has emerged as an effective solution for improving the sustainability of energy systems while reducing related costs. However, EDR suffers from capacity, proximity, latency, and reputation constraints, which complicate energy consumption in edge systems. To address these issues, this article proposes a novel approach for demand response in green MEC based on a contract system and edge energy harvesting (EEH). A time-varying hypergraph is proposed to formally describe an MEC scenario with time-varying features and multicommunity relationships. Informed by social contract theory, a contract system is developed for MEC based on the time-varying hypergraph. An EEH strategy based on the contract system is proposed to reduce the energy consumption of edge nodes. An edge demand response approach based on the EEH-enabled contract system (EDR-CS) is proposed to minimize edge utility (EU) model issues while preventing deceptions by edge nodes. The experimental results show that the proposed EDR-CS approach outperforms the existing baselines in terms of total completion time (TCT), task failure rate (TFR), overall data rate (ODR) and system energy consumption (SEC).

Synthetic aperture radar (SAR) is a high-resolution remote sensing technology widely employed for ground and sea surface target detection. However, due to the unique imaging mechanism and information representation of SAR images, conventional spatial-domain feature extraction methods often struggle to fully capture their discriminative features. To address this limitation, this letter introduces the wavelet domain as an additional feature extraction space and proposes a dual-domain feature-enhanced network based on wavelet attention for SAR ship detection. Specifically, two wavelet attention modules are designed to independently and jointly compute attention for high-frequency and low-frequency features in the wavelet domain. Meanwhile, an embedding grouping strategy is adopted to reduce computational costs while enhancing the model’s detailed perception and global understanding of ship targets. Furthermore, a dynamic domain fusion module is proposed to more effectively integrate wavelet-domain and spatial-domain information, enriching feature representation. Comprehensive experiments on two widely used SAR ship datasets demonstrate that the proposed method outperforms many other state-of-the-art detectors. The source code is available at https://github.com/Wenjing-Jiang-hbu/DFWA-Net.

Abstract Coronavirus disease 2019 (COVID-19) has swiftly spread throughout the globe, causing widespread infection in various countries and regions, and was declared a pandemic by World Health Organization (WHO) in 2020. Computer algorithms and models can help in the identification and classification of the COVID-19 virus in the medical domain, especially in CT, and X-rays and Electrocardiography tests with rapid and accurate results. In this paper, a COVID-19 electrocardiography classification model based on grey wolf optimization and support vector machine will be presented. A public online electrocardiography dataset was investigated in this paper with two classes (COVID-19, and Normal. The proposed model consists of three phases. The first phase is the feature extraction based on Resnet50. The second phase is the feature selection based on grey wolf optimization. The third phase is the classification based on the support vector machine. The experimental trials show that the proposed model achieves the highest accuracy possible when it is compared with other models that use different feature extraction and selection models, such as Alexnet and whale optimization algorithms. Also, the proposed model achieves the highest testing accuracy possible with 99.1% while related work that used hexaxial feature mapping and deep learning achieved 96.20% with an improvement of 2.9%. The achieved testing accuracy and its performance metrics such as Precision, Recall, and F1 Score support the research findings that the proposed model, while achieving the highest accuracy possible, it also consumes less time in the training by selecting a minimum number of features if it is compared with other related works which use the same dataset.

Micro-expressions (MEs) are tiny muscular movements on the face that conceal an individual's genuine emotions. However, the micro-expression recognition (MER) task faces challenges like short duration, low motion intensity, and a scarcity of training data. To solve these problems and obtain a good recognition effect, a Channel Self-Attention Residual Network (CSARNet) is proposed for extracting micro-expression discriminative information from motion stream images with augmented local features. Firstly, based on the offset frames of micro-expressions, a local feature augmentation strategy is devised to augment the local feature representations of motion flow images, thus effectively suppressing the interference of motions that are not related to micro-expressions. Second, aiming to mitigate the risk of model overfitting resulting from the dataset's limited size, CSARNet with a lightweight backbone network structure is designed to streamline the model's complexity and decrease computation time, which also accurately extracts the discriminative information of micro-expressions across channel and spatial dimensions, enabling the effective recognition of emotions. The proposed method was extensively tested on three benchmark datasets (SMIC, CASME II, SAMM) and the composite 3DB dataset, with experimental results clearly showcasing its superiority.

Despite the high spectral resolution and abundant information of hyperspectral images (HSI), their spatial resolution is relatively low due to limitations in sensor technology. Sensors often need to sacrifice some spatial resolution to ensure accurate light energy measurement when pursuing high spectral resolution. This trade-off results in HSI’s inability to capture fine spatial details, thereby limiting its application in scenarios requiring high-precision spatial information. HSI and multispectral images (MSI) fusion is a commonly used technique for generating high-resolution HSI (HR-HSI). However, many deep learning-based HSI-MSI fusion algorithms ignore correlation and multi-scale information between input images. To address this issue, we propose an asymptotic multi-scale symmetric fusion network (AMSF-Net) for hyperspectral and multispectral image fusion. AMSF-Net consists of two parts: the multi-level feature fusion (MFF) module and the progressive cross-scale spatial perception (PCP) module. The MFF module uses multi-stream feature extraction branches to perform information interaction between HSI and MSI at the same scale layer by layer, compensating for the spatial details lacking in HSI and the spectral details absent in MSI. The PCP module combines the input and output features of MFF, utilizes multi-scale bidirectional strip convolution and deep convolution to further refine edge features, and reconstructs HR-HSI by learning the features of different expansion roll branches by connecting across scales. Comparative experiments with several state-of-the-art HSI-MSI fusion algorithms on four publicly available datasets, CAVE, Chikusei, Houston and WorldView-3 are conducted to validate the effectiveness and superiority of AMSF-Net. On the Chikusei dataset, improvements were 9.1%, 12.5%, and 5.1%, respectively, on the indicators RMSE, ERGAS, and SAM, compared to the suboptimal method.

Motorimagery EEG classification plays a crucial role in non-invasive Brain-Computer Interface (BCI) research. However, the performance of classification is affected by the non-stationarity and individual variations of EEG signals. Simply pooling EEG data with different statistical distributions to train a classification model can severely degrade the generalization performance. To address this issue, the existing methods primarily focus on domain adaptation, which requires access to the test data during training. This is unrealistic and impractical in many EEG application scenarios. In this paper, we propose a novel multi-source domain generalization framework called EEG-DG, which leverages multiple source domains with different statistical distributions to build generalizable models on unseen target EEG data. We optimize both the marginal and conditional distributions to ensure the stability of the joint distribution across source domains and extend it to a multi-source domain generalization framework to achieve domain-invariant feature representation, thereby alleviating calibration efforts. Systematic experiments conducted on a simulative dataset, BCI competition IV 2a, 2b, and OpenBMI datasets, demonstrate the superiority and competitive performance of our proposed framework over other state-of-the-art methods. Specifically, EEG-DG achieves average classification accuracies of 81.79% and 87.12% on datasets IV-2a and IV-2b, respectively, and 78.37% and 76.94% for inter-session and inter-subject evaluations on dataset OpenBMI, which even outperforms some domain adaptation methods.

Mobile Edge Cache allows CDN edge nodes to be deployed closer to users, reducing content transmission latency in Vehicular Networks (VANETs). However, how to effectively utilize the limited storage space of cache nodes is the main issue in current research. To address this problem, we propose an incentive-driven hierarchical collaborative caching algorithm based on D3QN combined with user preferences (PC-ID3QN). Firstly, we constructed a UAV-assisted vehicular content-centric network framework, designing different collaborative caching strategy for various layers based on user preferences. Secondly, we proposed a hierarchical incentive mechanism based on social priority to motivate users to participate in collaborative sharing, thereby enhancing the utilization of system caching resources. Next, we modeled the cache placement problem as a system utility maximization problem and proved its NP-hardness. Then, by adjusting the constraint conditions, we conducted theoretical analysis and transform it into a linear programming(LP) problem to obtain an offline theoretical solution. This solution was validated against the simulation optimal solution obtained using the proposed PC-ID3QN algorithm, demonstrating the effectiveness of the algorithm. Finally, we validate the proposed caching strategy using the MovieLens dataset and conduct extensive experiments to verify the applicability and superiority of our solution in improving cache utility. Compared with DDQN, Dueling DQN and DQN, our proposed ID3QN algorithm reduces the request delay by 1.03%,1.73% and 2.21%, and reduces the energy cost by 38.8%, 19.6% and 17.2%, respectively.

To cope with the dramatic growth of future network traffic and the diversity of services, beyond fifth-generation (B5G) and sixth-generation (6G) wireless communication need to balance the network load while being service-oriented. In this paper, we consider a multi-access edge computing (MEC) driven radio access network (RAN) slicing scenario in a heterogeneous cellular network (HetNet) with local traffic overload. First, since users are generally non absolutely rational when selecting base stations (BSs), an evolutionary game (EG) based user association (UA) scheme is proposed to solve the problem of overload. Specifically, we innovatively define the load function of the base stations (BSs) and combine resource capacity of BSs to form the payoff function to dynamically adjust the load. Second, we model the network slicing (NS) problem using the transmission rate, average latency and quality of service (QoS). The problem is further relaxed and an NS algorithm based on distributed successive convex approximation (DSCA) is presented to derive a theoretical upper bound reference value as a static offline criterion. Finally, considering the high randomness of user task arrival in the real scenario, we formulate the multi-base station slicing problem as a stochastic game (SG) and a multi-agent twin delayed deep deterministic policy gradient (MATD3)-based distributed network slicing algorithm is proposed to obtain excellent slicing strategies. Simulation results show that our proposed UA algorithm has a unique evolutionary equilibrium (EE) solution and is highly scalable. The proposed MATD3-based NS algorithm has better performance compared to other baseline algorithms and converges to a utility value that best approximates the theoretical upper bound.

The rapid diagnosis of COVID-19 has become a pressing issue due to the strain the outbreak has placed on the healthcare system. This article aims to investigate the rapid and accurate diagnosis of COVID-19. This paper first introduces several widely used COVID-19 diagnostic techniques: rRT-PCR has excellent specificity and sensitivity, making it one of the most trustworthy ways to find the SARS-CoV-2 virus. Diagnostics based on X-rays are frequently employed as an adjunctive method. CT-based diagnosis can offer comprehensive details regarding lung health. It then highlights how machine learning combined with X-ray and CT images can be used to diagnose COVID-19. This approach can improve the accuracy and efficiency of detecting and evaluating the disease, helping healthcare professionals make decisions. Several standard machine learning methods are introduced, including supervised, unsupervised, and semi-supervised learning. Lastly, it forecasts machine learning development in the healthcare sector.

Objectives: Accurate extraction of regions of interest (ROI) with variable shapes and scales is one of the primary challenges in medical image segmentation . Current U-based networks mostly aggregate multi-stage encoding outputs as an improved multi-scale skip connection. Although this design has been proven to provide scale diversity and contextual integrity, there remain several intuitive limits: (i) the encoding outputs are resampled to the same size simply, which destruct the fine-grained information. The advantages of utilization of multiple scales are insufficient. (ii) Certain redundant information proportional to the feature dimension size is introduced and causes multi-stage interference. And (iii) the precision of information delivery relies on the up-sampling and down-sampling layers, but guidance on maintaining consistency in feature locations and trends between them is lacking. Methods: To improve these situations, this paper proposed a U-based CNN network named HAD-Net, by assembling a new hyper-scale shifted aggregating module (HSAM) paradigm and progressive reusing attention (PRA) for skip connections, as well as employing a novel pair of dual-branch parameter-free sampling layers, i.e. max-diagonal pooling (MDP) and max-diagonal un-pooling (MDUP). That is, the aggregating scheme additionally combines five subregions with certain offsets in the shallower stage. Since the lower scale-down ratios of subregions enrich scales and fine-grain context. Then, the attention scheme contains a partial-to-global channel attention (PGCA) and a multi-scale reusing spatial attention (MRSA), it builds reusing connections internally and adjusts the focus on more useful dimensions. Finally, MDP and MDUP are explored in pairs to improve texture delivery and feature consistency, enhancing information retention and avoiding positional confusion. Results: Compared to state-of-the-art networks, HAD-Net has achieved comparable and even better performances with Dice of 90.13%, 81.51%, and 75.43% for each class on BraTS20, 89.59% Dice and 98.56% AUC on Kvasir-SEG, as well as 82.17% Dice and 98.05% AUC on DRIVE. Conclusions: The scheme of HSAM+PRA+MDP+MDUP has been proven to be a remarkable improvement and leaves room for further research.

This paper presents an improved version of the U-Net segmentation framework for medical image segmentation, called CHWS-UNet. To build the proposed framework CHWS-UNet, we first develop a novel lightweight channel attention module called LCAM, based on which we further propose the Channel-Height-Width-Spatial (CHWS) attention module for channel, height, width, and spatial dimension-level feature refinement. Our CHWS-UNet is constructed by integrating the proposed CHWS attention modules into the shortcut paths between the encoder and the decoder stem. To justify the effectiveness of the proposed modules and networks, we then carried out extensive experiments on four public medical image datasets, including BUSI, ISIC2017, ISIC2018, PH and a proprietary uterus lesion ultrasound dataset from Shenzhen Maternity and Child Healthcare Hospital. The results show that the proposed attention module can significantly improve the performance of baseline models, even on small medical image datasets, without introducing noticeable parameters and computational costs. Further, the proposed segmentation framework can achieve promising performance compared to edge-cutting frameworks. The code can be found at CHWS-UNet.

The rapid proliferation of medical imaging technologies presents a significant challenge for cross-domain adaptive image detection, as lesion representations can vary dramatically across technologies. To address this issue, we draw inspiration from large language models to propose EAFP-Med, an efficient adaptive feature processing module based on prompts for medical image detection. EAFP-Med incorporates a prompt-driven dynamic parameter update mechanism, empowering it to extract cross-domain multi-scale lesion features from medical images of diverse modalities adaptively. This exceptional flexibility liberates it from the constraints of any particular imaging technique, fostering great adaptability. Furthermore, EAFP-Med can also serve as a feature preprocessing module connected to any model front-end to enhance the lesion features in input images. Moreover, we propose a novel adaptive disease detection model named EAFP-Med ST, which utilizes the Swin Transformer V2 – Tiny (SwinV2-T) as its backbone and connects it to EAFP-Med. We have compared our method to nine state-of-the-art methods. Experimental results show that the overall accuracy of EAFP Med ST on chest X-ray, brain magnetic resonance imaging, and skin image datasets is 98.47%, 97.60%, and 99.06%, respectively, superior to all the compared state-of-the-art methods.

High dimensional and small samples characterize gene expression data and contain a large number of genes unrelated to disease. Feature selection improves the efficiency of disease diagnosis by selecting a small number of important genes. Unfortunately, existing algorithms do not consider the correlation between features, and search algorithms tend to fall into the local optimal solution in the feature search process. To this end, this paper proposes a feature filter and group evolution hybrid feature selection algorithm (FG-HFS) for high-dimensional gene expression data. Unlike existing algorithms, we propose using spectral clustering to group redundant features into a group. Then, we propose a redundant feature filter algorithm. According to the principle of approximate Markov blanket, grouped feature groups are filtered to delete these redundant features. Among them, filtered features are evenly divided by density according to the feature exponential strategy. Most importantly, we propose using the group evolution multi-objective genetic algorithm to search the filtered feature subsets and evaluate the candidate feature subsets according to the in-group and out-group so as to select the feature subsets with the highest accuracy and the least number. Experimental results show that the average accuracy (ACC) and Matthews correlation coefficient (MCC) indexes of the selected feature subsets (FSs) by the FG-HFS algorithm on 5 gene expression datasets are 92.76% and 88.76%, respectively, which are significantly better than the existing algorithms. In addition, the FSs and ACC/FSs indexes of the FG-HFS algorithm are also better than the existing algorithms, which fully proves the superiority of the FG-HFS algorithm. More importantly, the Wilcoxon and Friedman statistical experiments results show that the feature selection effect of FG-HFS algorithm is significantly better than that of existing algorithms, no matter in pairwise comparison or multiple comparison.

Abstract X‐ray security checks aim to detect contraband in luggage; however, the detection accuracy is hindered by the overlapping and significant size differences of objects in X‐ray images. To address these challenges, the authors introduce a novel network model named Multi‐Scale Feature Attention (MSFA)‐DEtection TRansformer (DETR). Firstly, the pyramid feature extraction structure is embedded into the self‐attention module, referred to as the MSFA. Leveraging the MSFA module, MSFA‐DETR extracts multi‐scale feature information and amalgamates them into high‐level semantic features. Subsequently, these features are synergised through attention mechanisms to capture correlations between global information and multi‐scale features. MSFA significantly bolsters the model's robustness across different sizes, thereby enhancing detection accuracy. Simultaneously, A new initialisation method for object queries is proposed. The authors’ foreground sequence extraction (FSE) module extracts key feature sequences from feature maps, serving as prior knowledge for object queries. FSE expedites the convergence of the DETR model and elevates detection accuracy. Extensive experimentation validates that this proposed model surpasses state‐of‐the‐art methods on the CLCXray and PIDray datasets.

Integrating Internet of Things (IoT) technologies in the healthcare industry represents a transformative shift with tangible benefits. This paper provides a detailed examination of IoT adoption in healthcare, focusing on specific sensor types and communication methods. It underscores successful real-world applications, including remote patient monitoring, individualized treatment strategies, and streamlined healthcare delivery. Furthermore, it delves into the intricate challenges to realizing the full potential of IoT in healthcare. This includes addressing data security concerns, ensuring seamless interoperability, and optimizing the use of IoT-generated data. The paper seeks to inspire practitioners and researchers by highlighting the practical implications of IoT in healthcare, emphasizing the ways IoT can enhance patient care, resource allocation, and overall healthcare efficiency.

Brains are the control center of the nervous system in human bodies, and brain tumor is one of the most deadly diseases. Currently, magnetic resonance imaging (MRI) is the most effective way to brain tumors early detection in clinical diagnoses due to its superior imaging quality for soft tissues. Manual analysis of brain MRI is error-prone which depends on empirical experience and the fatigue state of the radiologists to a large extent. Computer-aided diagnosis (CAD) systems are becoming more and more impactful because they can provide accurate prediction results based on medical images with advanced techniques from computer vision. Therefore, a novel CAD method for brain tumor classification named RanMerFormer is presented in this paper. A pre-trained vision transformer is used as the backbone model. Then, a merging mechanism is proposed to remove the redundant tokens in the vision transformer, which improves computing efficiency substantially. Finally, a randomized vector functional-link serves as the head in the proposed RanMerFormer, which can be trained swiftly. All the simulation results are obtained from two public benchmark datasets, which reveal that the proposed RanMerFormer can achieve state-of-the-art performance for brain tumor classification. The trained RanMerFormer can be applied in real-world scenarios to assist in brain tumor diagnosis.

With the rapid development of applications with different use cases and service demands for edge network, network slicing is an emerging solution for satisfying service-oriented requirements, while the low earth orbit (LEO) satellite caching-assisted communication has been considered as one of the key elements for effective services. With limited resources at the edge of the radio access network (RAN), it is challenging to take advantage of the LEO content cache to joint allocation of communication, computing and caching space (3C) resources. To this end, we investigate the problem of resource slicing and scheduling of joint 3C resources in RAN edge scenario assisted by LEO content caching. A hierarchical resource slicing framework is proposed for dynamic allocation of multidimensional resources. The optimization variables are relaxed and the constraints are adjusted. The sequential quadratic programming (SQP) iteration algorithm is proposed as theoretical offline baseline. Due to its complex solving process and limited real-time performance, we incorporate Long Short-Term Memory (LSTM) into the Soft Actor-Critic (SAC) algorithm to aware extract the distribution characteristics of historical information and propose the deep reinforcement learning algorithm of LSTM-SAC. Meanwhile, the proportional priority based scheduling algorithm is employed in the intra-slice. Compared to SAC, TD3 and DDPG algorithms, the proposed algorithm is the closest to the theoretical value, improves the objective function by 6.95%, 9.52% and 11.52% respectively, which can significantly improve the system rate while satisfying the service level agreements.

Abstract Amongst all types of cancer, breast cancer has become one of the most common cancers in the UK threatening millions of people’s health. Early detection of breast cancer plays a key role in timely treatment for morbidity reduction. Compared to biopsy, which takes tissues from the lesion for further analysis, image-based methods are less time-consuming and pain-free though they are hampered by lower accuracy due to high false positivity rates. Nevertheless, mammography has become a standard screening method due to its high efficiency and low cost with promising performance. Breast mass, as the most palpable symptom of breast cancer, has received wide attention from the community. As a result, the past decades have witnessed the speeding development of computer-aided systems that are aimed at providing radiologists with useful tools for breast mass analysis based on mammograms. However, the main issues of these systems include low accuracy and require enough computational power on a large scale of datasets. To solve these issues, we developed a novel breast mass classification system called DF-dRVFL. On the public dataset DDSM with more than 3500 images, our best model based on deep random vector functional link network showed promising results through five-cross validation with an averaged AUC of 0.93 and an average accuracy of $$81.71\%$$ 81.71 % . Compared to sole deep learning based methods, average accuracy has increased by 0.38. Compared with the state-of-the-art methods, our method showed better performance considering the number of images for evaluation and the overall accuracy.

Alzheimer's disease (AD) is the leading neurodegenerative disorder and primary cause of dementia. Researchers are increasingly drawn to automated diagnosis of AD using neuroimaging analyses. Conventional deep learning (DL) models excel in constructing learning classifiers in early-stage AD diagnosis. However, they often struggle with AD diagnosis due to uncertainties stemming from unclear annotations by experts, challenges in data collection, such as data harmonization issues, and limitations in equipment resolution. These factors contribute to imprecise data, hindering accurate analysis, interpretation of obtained results, and understanding of complex symptoms. In response, the integration of fuzzy logic into DL, forming fuzzy deep learning (FDL), effectively manages imprecise data and provides interpretable insights, offering a valuable advancement in AD. Therefore, exploring recent advancements in integrating DL with fuzzy logic is crucial for improving AD diagnosis. In this review, we explore the contributions of fuzzy logic within FDL models, focusing on fuzzy-based image preprocessing, segmentation, and classification. Moreover, in exploring research directions, we discuss the possibility of the fusion of multimodal data with fuzzy logic, addressing challenges in AD diagnosis. Leveraging fuzzy logic and membership while integrating diverse datasets, such as genomics, proteomics, and metabolomics may provide an effective development of a DL classifier. In addition, fuzzy explainable DL promises more accurate and linguistically interpretable decision support systems for AD diagnosis. The primary objective of this article is to serve as a comprehensive and authoritative resource for newcomers, researchers, and clinicians interested in employing FDL models for AD diagnosis.

Synthetic aperture radar (SAR) tends to be seriously affected by speckle noise due to its inherent imaging characteristics, which brings great challenges to the high-level visualization task of SAR images. Therefore, speckle suppression plays a crucial role in remote sensing image processing. Attention-based SAR image denoising algorithms frequently struggle to capture rich feature information and face challenges in balancing the trade-off between denoising and preserving texture details. To solve the above problems, this paper constructs a local and global dual-branch network (LG-DBNet) for SAR image denoising. This network can effectively suppress speckle noise while fully retaining the detail information of the original image. Firstly, the shallow features are extracted through simple convolution. Then, a dual-branch structure constructed using different attention modules is used to extract deep features from SAR images. Specifically, one branch performs local deep feature extraction of an image through a hybrid attention module built by a convolutional neural network (CNN), while the other branch utilizes a superposition of self-attention mechanisms for global deep feature extraction of the image. Finally, the final denoised image is generated through global residual learning. LG-DBNet can effectively extract the local and global image information through the dual-branch structure, and further focus on the noise information, which can better retain the texture information of the image while effectively denoising. The experimental results show that compared with the state-of-the-art SAR image denoising algorithms, the proposed algorithm not only improves on various objective indexes, but also shows great advantages in the visual effect after denoising.

With the development of intelligent transportation, growing attention has been received to integrated sensing and communication (ISAC) systems. In this paper, we formulate a novel passive sensing technique to obtain information on the vehicle’s direction of arrival (DOA) using reconfigurable intelligent surfaces (RIS). A novel estimation method is proposed in the scenario with a receiver using only one full-functional channel, where multiple measurements for the DOA estimation are achieved by controlling the reflection matrix (measurement matrix) in the RIS. Moreover, different from the existing estimation methods, we also consider the interference signals introduced by wireless communication in the ISAC system. Then, we propose a novel atomic norm-based method to remove the interference signals and reconstruct the sparse signal. Additionally, a novel Hankel-based multiple signal classification (MUSIC) method is formulated to obtain the DOA information after the interference removal. To reduce the interference signals more efficiently and improve the performance of the sparse reconstruction, we optimize the measurement matrix to improve the signal-to-interference-plus-noise ratio (SINR). Finally, the theoretical Cram’er-Rao lower bound (CRLB) is derived for the ISAC system on the vehicle DOA estimation. Simulation results show that the proposed method can achieve better performance in the DOA estimation, and the corresponding CRLB with different distributions of the sensing nodes are shown. The code for the proposed method is available online https://github.com/chenpengseu/PassiveDOA-ISAC-RIS.git.

With the development of Intelligent Transportation Systems (ITS), edge caching has gradually emerged as a critical technology to reduce transmission delay and optimize network load. However, the limited storage capacity and service scope of individual cache servers significantly degrade the performance of edge caching. To address this issue, we propose a social-aware assisted edge collaborative caching algorithm based on Dueling Double Deep Q-Network and Digital Twin Network (SACTD-D3). The algorithm can dynamically adjust the caching decision based on the similarity of user semantic information and the availability of edge services to fully utilize the caching capacity of edge servers. Firstly, vehicle clusters are formed based on users’ semantic similarity, and an on-board cloud is constructed to reduce user request delay by sinking edge services. Secondly, based on the establishment of the three-layer structure of macro base station, roadside units and on-board cloud, the content heat-based caching decision policy is utilized to effectively improve the content cache hit rate. Moreover, an optimization problem is formulated to maximize the overall utility of the system subject to transmission delay and system cost, and thus the optimal solution is obtained using the proposed$\varepsilon$-greedy SACTD-D3 algorithm. Furthermore, due to the dynamic complexity of the network topology, digital twin is used to simplify and map the network topology into digital twin networks for analysis and processing to improve network efficiency. Finally, the simulation results demonstrate the effectiveness of the proposed algorithm in improving the system performance. Compared with Double DQN, Dueling DQN and DQN, the proposed SACTD-D3 algorithm reduces the request delay by 2.62$\%$, 3.06$\%$and 3.95$\%$, and reduces the energy cost by 26.07$\%$, 47.05$\%$and 49.90$\%$, respectively.

To improve reconstruction performance in imagery compressive sensing, the present paper changes solving a block image compressive sensing reconstruction into a convex optimization problem. First, a Total-Variation norm minimization constraints model that includes both L1 and L2 norm functions is established. The split Bregman iterative method solves the model with convex optimization. Then, a robust adaptive image block compressive sensing algorithm is studied based on an analysis of the image features. The image is divided into blocks, and an overlap image block compressive reconstruction method is proposed. Finally, to solve the block effect caused by block compressive sensing reconstruction, a novel image overlap block compressive sensing reconstruction based on the Poisson function is suggested to avoid the block effect in the reconstruction process. The experimental results show that compared with other traditional compressive sensing reconstruction algorithms, the proposed method can generate a better image reconstruction result. According to the PSNR evaluation, when the sampling rate is 0.3, the proposed method is improved by more than 20.98% compared to the conventional techniques, and according to the SSIM evaluation, it has improved by more than 11.92% from the traditional methods. We can also find that the proposed method has better construction effect for traffic sign image recognition compared with ordinary natural image reconstruction. When the sampling rate is only 0.1, the PSNR value reaches 44.28dB, and the SSIM reconstruction accuracy reaches 98.14%. After reconstructing different types and characteristic images, it is supported that the proposed algorithm has good robustness and anti-noise performance.

Existing deep learning-based wireless sensing models usually require intensive computation. In this paper, we introduce a lightweight RF-based 3D human pose estimation model, i.e., Fast RFPose, to enable real-time human pose estimation. Specifically, Fast RFPose first estimates the human locations in the RF heatmap and crops the human location regions, then estimates the fine-grained human poses based on the cropped small RF heatmaps. In the experiments, we build a radio system and a multi-view camera system to acquire the RF signals and the ground-truth human poses, and compare Fast RFPose with state-of-the-art methods. Experimental results demonstrate that Fast RFPose outperforms the alternative methods. Besides, we further deploy the trained Fast RFPose model on a laptop with a CPU and Fast RFPose can achieve 66 FPS processing speed, which means it can meet the real-time running requirements in mobile devices.

For learning algorithms, accessing large volumes of annotated data is highly desirable but not always available, especially in real-world scenarios. Accordingly, learning in the high-dimensional and low-sample size (HDLS) domain is recognised as one of the core challenges for modern AI systems. In this work, we consider a particular but very practical scenario in the HDLS domain where the number of training samples is not limited to mere few observations but yet it is not large enough to reliably build models with high degrees of expressivity. To address the problem, we present a new neuromorphic algorithm capable of fine-tuning existing feature spaces via learning relevant associations in high dimensional data with high probability. The algorithm is based on the idea of Concept Cells [1] and mimics properties attributed to memory and learning inherent to live neural systems. We demonstrate, through numerous numerical experiments, that the algorithm can “fine-tune” and “adapt” the feature space of pre-trained neural networks for better performance on new tasks in the HDLS domain. In addition, we study the impact of this “tuning” on quasi-orthogonal measures, which correlates with classification and calibration metrics.

Abstract The rampant spread of explicit content across social media can leave a damaging mark on our society. Hence, the need to be vigilant in detecting and curtailing sexually explicit content cannot be overstated. As such, it becomes paramount to discern and manage sexually explicit material to curb its dissemination and safeguard our digital communities from its harmful effects. In this article, we propose a unique technique entitled attention‐enabled pooling (ABP) embedded Swin transformer‐based YOLOv3 (ASYv3) for the detection of obscene areas present in the images with a bounding box around the offensive regions. ASYv3 employs a unique two‐step approach for enhanced performance in obscene detection. In the first step, a scalable and efficient Swin transformer block is integrated, utilizing self‐attention and model parallelism to train massive models effectively. In the second phase, the embedding layer of the Swin transformer is replaced with ABP, mitigating disruption of feature context. ABP allows for the projection of raw‐valued features into linear form with proper attention to feature context information at specified locations, resulting in optimized feature extraction. The proposed ABP embedded Swin transformer‐based YOLOv3 (ASYv3) was trained with annotated obscene images (AOI) dataset. The proposed ASYv3 model surpassed the state‐of‐the‐art methods by achieving 97% testing accuracy, 96.62% precision, 97.40% sensitivity, 3.48% FPR rate, 97.37% NPV values, and 95.59% mAP values, respectively.

Abstract Countless cybercrime instances have shown the need for detecting and blocking obscene material from social media sites. Deep learning methods (DLMs) outperformed in recognizing obscene content flooded on many online platforms. However, these contemporary DLMs primarily treat the recognition of obscene content as a simple task of binary classification, rather than focusing on the labelling of obscene areas. Hence, many of these methods could not pay attention to the fact that misclassification samples are so diverse. Therefore, this paper focuses on two aspects (i) developing a deep learning model that could classify and label the obscene portion, and (ii) generating a labelled obscene image dataset with a wide variety of obscene samples to minimize the risks of inaccurate recognition. We have proposed a method named S3Pooling based bottleneck attention module (BAM) embedded MobileNetV2‐YOLOv3 (SBMYv3) for automatic detection of obscene content using an attention mechanism and a suitable pooling strategy. The key contributions of our article are: (i) generation of a well‐labelled obscene image dataset with a variety of augmentation strategies using Pix‐2‐Pix GAN (ii) modifications to the backend architecture of YOLOv3 using MobileNetV2 and BAM to ensure focused and accurate feature extraction, and (iii) selection of an optimal pooling strategy, that is, S3Pooling strategy, while taking the design of the feature extractor into account. The proposed SBMYv3 model outperformed other state‐of‐the‐art models with 99.26% testing accuracy, 99.39% recall, 99.13% precision, and 99.13% IoU values respectively.

Abstract Facial expression recognition (FER) is widely used in many fields. To further improve the accuracy of FER, this paper proposes a method based on double‐code LBP‐layer spatial‐attention network (DLSANet). The backbone model for the DLSANet is an emotion network (ENet), which is modified with a double‐code LBP (DLBP) layer and a spatial attention module. The DLBP layer is at the front of the first convolutional layer. More valuable features can be extracted by inputting the image processed by DLBP into convolutional layers. The JAFFE and CK+ datasets are used, which contain seven expressions: happiness, anger, disgust, neutral, fear, sadness, and surprise. The average of fivefold cross‐validation shows that DLSANet achieves a recognition accuracy of 93.81% and 98.68% on the JAFFE and CK+ datasets. The experiment reveals that the DLSANet can produce better classification results than state‐of‐the‐art methods.

Abstract Automatic brain tumour segmentation in MRI scans aims to separate the brain tumour's endoscopic core, edema, non‐enhancing tumour core, peritumoral edema, and enhancing tumour core from three‐dimensional MR voxels. Due to the wide range of brain tumour intensity, shape, location, and size, it is challenging to segment these regions automatically. UNet is the prime three‐dimensional CNN network performance source for medical imaging applications like brain tumour segmentation. This research proposes a context aware 3D ARDUNet (Attentional Residual Dropout UNet) network, a modified version of UNet to take advantage of the ResNet and soft attention. A novel residual dropout block (RDB) is implemented in the analytical encoder path to replace traditional UNet convolutional blocks to extract more contextual information. A unique Attentional Residual Dropout Block (ARDB) in the decoder path utilizes skip connections and attention gates to retrieve local and global contextual information. The attention gate enabled the Network to focus on the relevant part of the input image and suppress irrelevant details. Finally, the proposed Network assessed BRATS2018, BRATS2019, and BRATS2020 to some best‐in‐class segmentation approaches. The proposed Network achieved dice scores of 0.90, 0.92, and 0.93 for the whole tumour. On BRATS2018, BRATS2019, and BRATS2020, tumour core is 0.90, 0.92, 0.93, and enhancing tumour is 0.92, 0.93, 0.94.

Abstract In this paper, a new sparsity‐optimised Farrow structure variable fractional delay (SFS‐VFD) filter is proposed to address the aperture effect in wideband array. Our method is based on coefficient (anti‐)symmetry and optimises the number and orders of its sub‐filters, greatly reducing the non‐zero coefficients. The established cost function is formulated as a parametric minimisation problem with multiple regularisation constraints, and solved by the modified three‐block alternating direction multiplier method (MTB‐ADMM), which is improved by introducing core variable correction items to ensure stable and fast convergence. Experimental results show that the SFS‐VFD filter reduces the complexity of the system by decreasing the use of multipliers and adders while ensuring high delay accuracy. In wideband array, the SFS‐VFD filter effectively corrects the aperture effect and achieves precise beam pointing.

Deep learning has become a primary choice in medical image analysis due to its powerful representation capability. However, most existing deep learning models designed for medical image classification can only perform well on a specific disease. The performance drops dramatically when it comes to other diseases. Generalizability remains a challenging problem. In this paper, we propose an evolutionary attention-based network (EDCA-Net), which is an effective and robust network for medical image classification tasks. To extract task-related features from a given medical dataset, we first propose the densely connected attentional network (DCA-Net) where feature maps are automatically channel-wise weighted, and the dense connectivity pattern is introduced to improve the efficiency of information flow. To improve the model capability and generalizability, we introduce two types of evolution: intra- and inter-evolution. The intra-evolution optimizes the weights of DCA-Net, while the inter-evolution allows two instances of DCA-Net to exchange training experience during training. The evolutionary DCA-Net is referred to as EDCA-Net. The EDCA-Net is evaluated on four publicly accessible medical datasets of different diseases. Experiments showed that the EDCA-Net outperforms the state-of-the-art methods on three datasets and achieves comparable performance on the last dataset, demonstrating good generalizability for medical image classification.

Oral diseases have a significant impact on human health, often going unnoticed in their early stages. Deep learning, a promising field in artificial intelligence, has shown remarkable success in various domains, especially dentistry. This paper aims to provide an overview of recent research on deep learning applications in dentistry, with a focus on dental imaging. Deep learning algorithms perform well in difficult tasks such as image segmentation and recognition, enabling accurate identification of oral conditions and abnormalities. Integration of deep learning with other oral health data offers a holistic understanding of the relationship between oral and systemic health. However, there are still many challenges that need to be addressed.

Internet of Medical Things (IoMT) refers to applying Internet of Things (IoT) into the medical field. The IoMT enables a medical system to connect various smart devices, such as wearable sensors, medical examination instruments, and hospital assets, for establishing an information platform. These smart devices act as the basic nodes in an IoMT system, collecting or generating health data and transmitting the data to the server for further processing and analysis. Physicians apply health data to make better medical decisions. Recently, the IoMT has been widely applied in many areas, including smart hospital, remote health monitoring, disease diagnosis, and infectious disease tracking. In this review, we investigated the IoMT from its concept and theory to its deployment domains, adopted technologies, and applications. We provided theoretical explanations with various examples and more than one hundred representative references. We also presented a cutting-edge discussion for the challenges and directions of the IoMT. We hoped that this systematic review would be beneficial to readers of all levels and backgrounds, including industry beginners, medical institution administrators, policy makers, and experienced researchers.

Inspired by the biological evolution, this paper proposes an evolutionary synthesis mechanism to automatically evolve DenseNet towards high sparsity and efficiency for medical image classification. Unlike traditional automatic design methods, this mechanism generates a sparser offspring in each generation based on its previous trained ancestor. Concretely, we use a synaptic model to mimic biological evolution in the asexual reproduction. Each generation’s knowledge is passed down to its descendant, and an environmental constraint limits the size of the descendant evolutionary DenseNet, moving the evolution process towards high sparsity. Additionally, to address the limitation of ensemble learning that requires multiple base networks to make decisions, we propose an evolution-based ensemble learning mechanism. It utilises the evolutionary synthesis scheme to generate highly sparse descendant networks, which can be used as base networks to perform ensemble learning in inference. This is specially useful in the extreme case when there is only a single network. Finally, we propose the MEEDNets (Medical Image Classification via Ensemble Bio-inspired Evolutionary DenseNets) model which consists of multiple evolutionary DenseNet-121s synthesised in the evolution process. Experimental results show that our bio-inspired evolutionary DenseNets are able to drop less important structures and compensate for the increasingly sparse architecture. In addition, our proposed MEEDNets model outperforms the state-of-the-art methods on two publicly accessible medical image datasets. All source code of this study is available at https://github.com/hengdezhu/MEEDNets.

Deep learning (DL) has succeeded in the field of target detection and has been introduced into the researches of ground moving target indication (GMTI) for synthetic aperture radar (SAR) recently. Due to the lack of labeled data in SAR/GMTI, simulated data are usually employed to support the training of networks, which has proved to be a feasible way in practice. Although some simulated data are very close to the real radar data, the fact is that distribution differences between them are inevitable and always lead to a performance loss of the network. Motivated by recent advances in transfer learning, this letter proposes a new method for ground moving target detection of multichannel SAR systems, namely, subdomain adaptive residual network (SARN). It is built on the basis of ResNet18, and subdomain adaptation is introduced. During the network training, multi-kernel local maximum mean discrepancy (MK-LMMD) is minimized as well as classification error. Experiments on three-channel SAR data show that the proposed method significantly improves the detection performance as compared with CA-CFAR and the DL method.

Due to buildings blocking GPS and Wi-Fi signals, traditional techniques can't offer the user's required positioning accuracy in resource-constrained underground parking, but the cooperation of agent nodes can provide the exact localization information to improve the positioning accuracy. However, some well-localized agents may not be willing to sacrifice additional power to improve the others' positioning accuracy. To encourage cooperation among nodes and allocate transmission power reasonably, this paper proposes a bidding-auction-based cooperative localization (BACL) algorithm to improve the positioning accuracy of agent nodes by joint node selection incentive and power allocation strategy. Firstly, the contribution of channel parameters and prior localization information of agent nodes for positioning accuracy are quantified and an incentive mechanism of cooperative localization from an economic perspective is proposed. Secondly, a virtual currency incentive rule is developed to compensate agent nodes of cooperative localization reasonably due to the consumption of energy for transmitting their location information. Finally, the simulation results have shown that the proposed BACL algorithm has excellent performance in terms of localization accuracy in resource-constrained scenarios. Compared with the full-power cooperative localization (FPCL) and non-cooperative localization (NCL) algorithms, the proposed BACL algorithm improved the positioning accuracy by 10% and 65%, respectively. Meanwhile, compared with the FPCL algorithm, the proposed algorithm reduced resource consumption by 50%.

With the continuous development of the Internet of Vehicles(IoV), Vehicle Edge Computing(VEC) has become a key technology for computational resource scheduling, but more and more smart devices are connected to the internet, which makes it difficult for traditional Vehicle Edge Networks(VEN) to deal with tasks in time. In this paper, in order to cope with the challenges of the large number of devices accessing the internet, we propose a vehicle-assisted computation offloading algorithm based on proximal policy optimization(VCOPPO) for User Equipment(UE) tasks, and it combines dynamic parked vehicles incentives mechanism and computational resource allocation strategy by using road vehicles and parked vehicles as edge servers. Firstly, a non-convex optimization problem combining VEN utility and task processing delay is formulated, subject to the constraints of the residual energy and the transmission rate of the task. Secondly, the proposed VCOPPO is used to solve the formulated non-convex optimization problem, and we use stochastic policy to obtain the optimal computation offloading decisions and resource allocation schemes. Finally, the experimental results have shown that the proposed VCOPPO has an excellent performance in network reward and task processing delay respectively, and it can effectively schedule and allocate computational resources. Compared with using Dueling Deep Q Network(Dueling DQN), Deep Q Network(DQN) and Q-learning methods, the proposed VCOPPO improves the network reward by 31%, 18% and 91%, reduces the delay in task processing by 78%, 63% and 74%, respectively.

In the global epidemic, distance learning occupies an increasingly important place in teaching and learning because of its great potential. This paper proposes a web-based app that includes a proposed 8-layered lightweight, customized convolutional neural network (LCCNN) for COVID-19 recognition. Five-channel data augmentation is proposed and used to help the model avoid overfitting. The LCCNN achieves an accuracy of 91.78%, which is higher than the other eight state-of-the-art methods. The results show that this web-based app provides a valuable diagnostic perspective on the patients and is an excellent way to facilitate medical education. Our LCCNN model is explainable for both radiologists and distance education users. Heat maps are generated where the lesions are clearly spotted. The LCCNN can detect from CT images the presence of lesions caused by COVID-19. This web-based app has a clear and simple interface, which is easy to use. With the help of this app, teachers can provide distance education and guide students clearly to understand the damage caused by COVID-19, which can increase interaction with students and stimulate their interest in learning.

Abstract COVID-19 has caused over 6.35 million deaths and over 555 million confirmed cases till 11/July/2022. It has caused a serious impact on individual health, social and economic activities, and other aspects. Based on the gray-level co-occurrence matrix (GLCM), a four-direction varying-distance GLCM (FDVD-GLCM) is presented. Afterward, a five-property feature set (FPFS) extracts features from FDVD-GLCM. An extreme learning machine (ELM) is used as the classifier to recognize COVID-19. Our model is finally dubbed FECNet. A multiple-way data augmentation method is utilized to boost the training sets. Ten runs of tenfold cross-validation show that this FECNet model achieves a sensitivity of 92.23 ± 2.14, a specificity of 93.18 ± 0.87, a precision of 93.12 ± 0.83, and an accuracy of 92.70 ± 1.13 for the first dataset, and a sensitivity of 92.19 ± 1.89, a specificity of 92.88 ± 1.23, a precision of 92.83 ± 1.22, and an accuracy of 92.53 ± 1.37 for the second dataset. We develop a mobile app integrating the FECNet model, and this web app is run on a cloud computing-based client–server modeled construction. This proposed FECNet and the corresponding mobile app effectively recognize COVID-19, and its performance is better than five state-of-the-art COVID-19 recognition models.

Deep ensemble learning, where we combine knowledge learned from multiple individual neural networks, has been widely adopted to improve the performance of neural networks in deep learning. This field can be encompassed by committee learning, which includes the construction of neural network cascades. This study focuses on the high-dimensional low-sample-size (HDLS) domain and introduces multiple instance ensemble (MIE) as a novel stacking method for ensembles and cascades. In this study, our proposed approach reformulates the ensemble learning process as a multiple-instance learning problem. We utilise the multiple-instance learning solution of pooling operations to associate feature representations of base neural networks into joint representations as a method of stacking. This study explores various attention mechanisms and proposes two novel committee learning strategies with MIE. In addition, we utilise the capability of MIE to generate pseudo-base neural networks to provide a proof-of-concept for a "growing" neural network cascade that is unbounded by the number of base neural networks. We have shown that our approach provides (1) a class of alternative ensemble methods that performs comparably with various stacking ensemble methods and (2) a novel method for the generation of high-performing "growing" cascades. The approach has also been verified across multiple HDLS datasets, achieving high performance for binary classification tasks in the low-sample size regime.

Many of the indigenous languages today are struggling to survive, and they are in danger of disappearing.Closely followed by Africa, Asia has the most indigenous languages.Though indigenous languages have many sources in existence from where we can obtain acceptable knowledge, mythology, history, and perception of their communities, their diversity is decreasing at an alarming rate due to social pressure, external forces, and demographic changes.The systematic disappearance of indigenous languages threatens the lives of millions of families, children, and indigenous communities as well as the survival of their languages worldwide.Most indigenous languages have no written form; which makes them difficult to process and analyze using computational models.However, these languages need to be preserved as they are rich in oral traditions, and they remain remarkably consistent and reliable over time.Presently, many researchers and scientists are actively finding more evidence on indigenous languages to create language processing models using a variety of techniques.Exploring more in terms of grammar, words, and unique rules of sound help us understand the language intuitively and create more efficient linguistic models.However, this process generally tends to be more complex as these languages have very few resources and are often spoken in remote areas by fewer people.Computational linguistics applies computer science techniques for the analysis and synthesis of written and spoken languages.The practical goal of using computational linguistics for indigenous languages is comprehensive.It helps formulate semantic and grammatical frameworks for distinguishing languages through the computationally manageable implementation of semantic and syntactic analysis.Hence, the discovery of more advanced computational linguistics processing algorithms and learning principles that can effectively use the structural and distributional properties of indigenous languages is crucial.It helps develop cognitively and neuroscientifically reasonable computational models that work in the same way that indigenous language processing and learning might occur in the brain.This special issue was dedicated to explain how computational linguistics and natural language processing algorithms make inferences and gain insights into existing data of low-resource indigenous languages.The content mainly focuses on innovative research that formalizes human communication and spoken indigenous languages into the computational system.We welcomed researchers and practitioners from industry and academia to present their contributions against this background.This special issue saw a total of 21 submissions, from which five papers were published.It was intentional to adhere to a strict acceptance rate and ensure that only the best papers in the scope of

Tuberculosis is a very deadly disease, with more than half of all tuberculosis cases dead in countries and regions with relatively poor health care resources. Fortunately, the disease is curable, and early diagnosis and medication can go a long way toward curing TB patients. Unfortunately, traditional methods of TB diagnosis rely on specialist doctors, which is lacking in areas with high TB mortality rates. Diagnostic methods based on artificial intelligence technology are one of the solutions to this problem. We propose a Deep Transferred EfficientNet with SVM (DTE-SVM), which replaces the pre-trained EfficientNet classification layer with an SVM classifier and achieves auspicious performance on a small dataset. After ten runs of 10-fold Cross-Validation, the DTE-SVM has a sensitivity of 93.89±1.96, a specificity of 95.35±1.31, a precision of 95.30±1.24, an accuracy of 94.62±1.00, and an F1-score of 94.62±1.00. In addition, our study conducted ablation studies on the effect of the SVM classifier on model performance and briefly discussed the results.

introduction Share on Special Section on “Advances in Cyber-Manufacturing: Architectures, Challenges, & Future Research Directions” Editors: Gautam Srivastava Brandon University, Canada Brandon University, CanadaSearch about this author , Jerry Chun-Wei Lin Silesian Uiversity of Technology, Poland Silesian Uiversity of Technology, PolandSearch about this author , Calton Pu Georgia Tech, USA Georgia Tech, USASearch about this author , Yudong Zhang University of Leicester, UK University of Leicester, UKSearch about this author Authors Info & Claims ACM Transactions on Internet TechnologyVolume 23Issue 4Article No.: 49pp 1–4https://doi.org/10.1145/3627990Published:17 November 2023Publication History 0citation0DownloadsMetricsTotal Citations0Total Downloads0Last 12 Months0Last 6 weeks0 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

Cell segmentation and counting is a time-consuming task and an important experimental step in traditional biomedical research. Many current counting methods require exact cell locations. However, there are few such cell datasets with detailed object coordinates. Most existing cell datasets only have the total number of cells and a global segmentation labelling. To make more effective use of existing datasets, we divided the cell counting task into cell number prediction and cell segmentation respectively. This paper proposed a lightweight fast multi-task multi-scale feature fusion model based on generative adversarial networks (MFGAN). To coordinate the learning of these two tasks, we proposed a Combined Hybrid Loss function (CH Loss) and used conditional GAN to train our network. We proposed a Lightweight Fast Multitask Generator (LFMG) which reduced the number of parameters by 20% compared with U-Net but got better performance on cell segmentation. We used multi-scale feature fusion technology to improve the quality of reconstructed segmentation images. In addition, we also proposed a Structure Fusion Discrimination (SFD) to refine the accuracy of the details of the features. Our method achieved non-Point-based counting that no longer needs to annotate the exact position of each cell in the image during the training and successfully achieved excellent results on cell counting and cell segmentation.

Quantum computing has made rapid progress in many fields, especially in artificial intelligence. Quantum machine learning is introduced to enhance the training rate of machine learning algorithms and hence, to achieve enhanced accuracy. The integration of quantum computing with artificial intelligence is incorporated with quantum neural networks and quantum tensor flow. Quantum neural networks play a vital role in enhancement of quantum computing progress in techniques of artificial intelligence. The aim of this special issue is to depict the importance and advances of quantum computing in a variety of fields, especially artificial intelligence. In the article by Jiao,1 fault tolerance for design of digital comparators is examined. A coplanar quantum-dot cellular automata (QCA) based fault tolerance comparator is proposed in order to achieve good simulation results. QCA designer needs 63 cells of QCA of 0.8μ2 and it requires three clock cycles delay to complete the task. In the article by Dogra et al.,2 an algorithm named multi-level filtering based image fusion algorithm is proposed for fusion of image information from MRI and CT scan images into a single vector. In the first step, source images are processed with the help of a filter of rolling guidance and then the calculation of the detail layer is performed. After that, guidance is used with the same rolling filter for the calculation of the base layer for edge preservation at a higher level. At the next step, in order to remove noise and other artifacts, all detail layers are fused together using the Karhunen Loeve transform and all the base layers are combined together by using the principle of weighted superimposition. In the article by Umer et al.,3 a method for early detection of COVID-19 is proposed in order to minimize the mortality rate. This method has two different phases: in the first phase, the pre-trained models like AlexNet and MobileNet are used and from their last fully connected layers deep features are extracted and then these extracted feature vectors are concatenated. Applying a method of feature selection, such as principal component analysis (PCA), the best features are then selected and then those features are passed to k-nearest neighbors algorithm (KNN) and support vector machine (SVM) for classification. In the second phase, a model of quantum transfer learning is used where a pre-trained model ResNet-18 is used for the collection of deep features and then these features are given to a four-qubit quantum circuit for the purpose of training with tuned hyper parameters. The proposed method is tested on two X-ray image datasets which are publicly available. In the article by Das et al.,4 the prediction of the OS period in a brain tumor is performed using the rain forest framework. Different algorithms such as eXtreme Gradient Boosting (XGBoost), light gradient boosting machine, and SVMs are used to take radiomic features and the efficiency of prediction is dependent upon the tumor volume, which is segmented after taking multiple MRI images. The complete tumor and its sub-parts are extracted from MRI model modalities by using the U-Net++ deep model and these are stacked together for the extraction of deep features via a convolutional neural network. In order to achieve increased accuracy, features are reduced by using PCA and then for the prediction of the OS period, the reduced radiomic feature set is used. The prediction results are tested for survival groups in both cases: for two classes and for three classes. Experiments are performed on the dataset of BraTs 2017 for both two classes and three classes by using different types of classifiers. In order to have an enhanced accuracy, different methods of optimization like PCA and genetic algorithm are used for the fusion of feature sets. In the article by Lepcha et al.,5 a piecewise regression model is proposed for learning the images through the mapping of resolution images from low to high. This is performed by utilizing filtering of the domain transform and an optimized framework of weighted least squares. First, the weighted least mean square is used to coarsen the given input images to extract multiscale information by building multiscale decompositions for edge preservations for which recursive filtering is also applied. The patterns named as Hadamard from low resolution images are collected for the classification of high resolution and low resolution image features. Finally, a piecewise linear regression model is used to learn the relationship of mapping between the classes of high resolution and low resolution image features.

Abstract Aim Fruit category classification is important in factory packing and transportation, price prediction, dietary intake, and so forth. Methods This study proposed a novel artificial intelligence system to classify fruit categories. First, 2D fractional Fourier entropy with rotation angle vector grid was used to extract features from fruit images. Afterwards, a five‐layer stacked sparse autoencoder was used as the classifier. Results Ten runs on the test set showed our method achieved a micro‐averaged F1 score of 95.08% for an 18‐category fruit dataset. Conclusion Our method gives better micro‐averaged F1 score than 10 state‐of‐the‐art approaches.

Abstract Human gait recognition (HGR) shows high importance in the area of video surveillance due to remote access and security threats. HGR is a technique commonly used for the identification of human style in daily life. However, many typical situations like change of clothes condition and variation in view angles degrade the system performance. Lately, different machine learning (ML) techniques have been introduced for video surveillance which gives promising results among which deep learning (DL) shows best performance in complex scenarios. In this article, an integrated framework is proposed for HGR using deep neural network and fuzzy entropy controlled skewness (FEcS) approach. The proposed technique works in two phases: In the first phase, deep convolutional neural network (DCNN) features are extracted by pre‐trained CNN models (VGG19 and AlexNet) and their information is mixed by parallel fusion approach. In the second phase, entropy and skewness vectors are calculated from fused feature vector (FV) to select best subsets of features by suggested FEcS approach. The best subsets of picked features are finally fed to multiple classifiers and finest one is chosen on the basis of accuracy value. The experiments were carried out on four well‐known datasets, namely, AVAMVG gait, CASIA A, B and C. The achieved accuracy of each dataset was 99.8, 99.7, 93.3 and 92.2%, respectively. Therefore, the obtained overall recognition results lead to conclude that the proposed system is very promising.

Special issue

Abstract Target enhancement is the most important task in a video surveillance system. In order to improve the accuracy and efficiency of target enhancement, and better deal with the subsequent recognition, tracking, behaviour understanding and other processing of targets, a deep learning‐based image enhancement algorithm for video surveillance scenes is proposed. First, the super‐resolution reconstruction of the image is carried out through the image super‐resolution reconstruction method based on the hybrid deep convolutional network to improve the sharpness of the image. Then, for the reconstructed video surveillance scene image, the watershed image enhancement algorithm based on morphology and region merging is used to realize the enhancement of the video surveillance scene image. Deep learning algorithms can improve the accuracy of image enhancement through iterative calculations. Experimental results show that after image enhancement in daytime, night and noisy video surveillance scenes, the maximum enhancement difference rate is less than 0.5%, the cross‐linking degree is close to 1, and the average image enhancement time is less than 1.3 s. It can realize image enhancement of video surveillance scenes and improve the image clarity of the video surveillance scene.

COVID-19 pneumonia started in December 2019 and caused large casualties and huge economic losses. In this study, we intended to develop a computer-aided diagnosis system based on artificial intelligence to automatically identify the COVID-19 in chest computed tomography images. We utilized transfer learning to obtain the image-level representation (ILR) based on the backbone deep convolutional neural network. Then, a novel neighboring aware representation (NAR) was proposed to exploit the neighboring relationships between the ILR vectors. To obtain the neighboring information in the feature space of the ILRs, an ILR graph was generated based on the k-nearest neighbors algorithm, in which the ILRs were linked with their k-nearest neighboring ILRs. Afterward, the NARs were computed by the fusion of the ILRs and the graph. On the basis of this representation, a novel end-to-end COVID-19 classification architecture called neighboring aware graph neural network (NAGNN) was proposed. The private and public data sets were used for evaluation in the experiments. Results revealed that our NAGNN outperformed all the 10 state-of-the-art methods in terms of generalization ability. Therefore, the proposed NAGNN is effective in detecting COVID-19, which can be used in clinical diagnosis.

In recent years, we witnessed a speeding development of deep learning in computer vision fields like categorization, detection, and semantic segmentation. Within several years after the emergence of AlexNet, the performance of deep neural networks has already surpassed human being experts in certain areas and showed great potential in applications such as medical image analysis. The development of automated breast cancer detection systems that integrate deep learning has received wide attention from the community. Breast cancer, a major killer of females that results in millions of deaths, can be controlled even be cured given that it is detected at an early stage with sophisticated systems. In this paper, we reviewed breast cancer diagnosis, detection, and segmentation computer-aided (CAD) systems based on state-of-the-art deep convolutional neural networks. The available data sets also indirectly determine CAD systems' performance, so we introduced and discussed the details of public data sets. The challenges remaining in CAD systems for breast cancer are discussed at the end of this paper. The highlights of this survey mainly come from three following aspects. First, we covered a wide range of the basics of breast cancer from imaging modalities to popular databases in the community; Second, we presented the key elements in deep learning to form the compactness for methods mentioned in reviewed papers; Third and lastly, the summative details in each reviewed paper are provided so that interested readers can have a refined version of these works without referring to original papers. Therefore, this systematic survey suits readers with varied backgrounds and will be beneficial to them.

The automation in the diagnosis of medical images is currently a challenging task. The use of Computer Aided Diagnosis (CAD) systems can be a powerful tool for clinicians, especially in situations when hospitals are overflowed. These tools are usually based on artificial intelligence (AI), a field that has been recently revolutionized by deep learning approaches. These alternatives usually obtain a large performance based on complex solutions, leading to a high computational cost and the need of having large databases. In this work, we propose a classification framework based on sparse coding. Images are first partitioned into different tiles, and a dictionary is built after applying PCA to these tiles. The original signals are then transformed as a linear combination of the elements of the dictionary. Then, they are reconstructed by iteratively deactivating the elements associated with each component. Classification is finally performed employing as features the subsequent reconstruction errors. Performance is evaluated in a real context where distinguishing between four different pathologies: control versus bacterial pneumonia versus viral pneumonia versus COVID-19. Our system differentiates between pneumonia patients and controls with an accuracy of 97.74%, whereas in the 4-class context the accuracy is 86.73%. The excellent results and the pioneering use of sparse coding in this scenario evidence that our proposal can assist clinicians when their workload is high.

The Internet of Things (IoT) is making the world around us smarter and more convenient. However, its extensive application has rendered security problems, such as the privacy of sensitive data, increasingly serious. Encryption provides an effective and important means of protecting data privacy in the IoT. Because of resource limitations, to achieve high efficiency IoT devices require an encryption scheme that ensures that the encryption phase does not incur a heavy data transmission overhead. By virtue of its many advantages, the use of public-key encryption in current applications is widespread. Identity-based public-key encryption (IBE) removes the obstacle raised by the sophisticated certificate management required by other schemes, and its efficiency renders it more suitable for application in the IoT. However, a problem that needs to be solved in IBE is the revocation of a user whose private key may have been exposed. In this article, we present an efficient and practical IBE scheme having a revocation functionality to preserve data privacy in IoT applications. Elements in the system, such as sensors and actuators, can exchange encrypted data directly or via a cloud server. If a private key is compromised, the private key generator can revoke its user. The security of our proposed scheme can be proved based on SM9 encryption and the bilinear Diffie–Hellman problem.

Abstract Extreme learning machine (ELM) is a training algorithm for single hidden layer feedforward neural network (SLFN), which converges much faster than traditional methods and yields promising performance. In this paper, we hope to present a comprehensive review on ELM. Firstly, we will focus on the theoretical analysis including universal approximation theory and generalization. Then, the various improvements are listed, which help ELM works better in terms of stability, efficiency, and accuracy. Because of its outstanding performance, ELM has been successfully applied in many real-time learning tasks for classification, clustering, and regression. Besides, we report the applications of ELM in medical imaging: MRI, CT, and mammogram. The controversies of ELM were also discussed in this paper. We aim to report these advances and find some future perspectives.

Semi-supervised multi-view learning has been an important research topic due to its capability to exploit complementary information from unlabeled multi-view data. This work proposes MMatch, a new semi-supervised discriminative representation learning method for multi-view classification. Unlike existing multi-view representation learning methods that seldom consider the negative impact caused by particular views with unclear classification structures (weak discriminative views). MMatch jointly learns view-specific representations and class probabilities of training data. The representations concatenated to integrate multiple views’ information to form a global representation. Moreover, MMatch performs the smoothness constraint on the class probabilities of the global representation to improve pseudo labels, whereas the pseudo labels regularize the structure of view-specific representations. A discriminative global representation is mined with the training process, and the negative impact of weak discriminative views is overcome. Besides, MMatch learns consistent classification while preserving diverse information from multiple views. Experiments on several multi-view datasets demonstrate the effectiveness of MMatch.

Deep learning-based detection methods have achieved great success in ship target detection in synthetic aperture radar (SAR) images. However, due to the interference of imaging mechanism, speckle noise, and sea and land clutter, ship detection in SAR images still suffers from difficult interpretation. It is found that most ship detection algorithms focus on object-level detection while ignoring pixel-level information. In order to further improve the recognition effectiveness and positioning accuracy of ships in SAR images, we present a novel ship detection method based on a feature interaction network (FINet) in SAR images from the perspective of object-level and pixel-level. FINet consists of an object-level detection network and a pixel-level detection network. The information of the two branches is fused through the feature interaction module (FIM), and then the object-level information and pixel-level information are enhanced by the feature guidance module (FGM). Finally, FINet utilizes object-level and pixel-level detection heads for prediction and regression to obtain object-level classification accuracy, positioning bounding box coordinates, and pixel-level binary classification results. The experimental results demonstrate that the classification effectiveness and localization accuracy of FINet are better than those of the comparison algorithms, and FINet achieves the best performance.

Synthetic aperture radar (SAR), due to its inherent characteristics, will produce speckle noise, which results in the deterioration of image quality, so the removal of speckle in SAR image is very important for the subsequent high-level image processing. In order to balance the relationship between denoising and texture preservation, we propose a multiscale residual dense dual attention network (MRDDANet) for SAR image denoising. This algorithm can effectively suppress the speckle while fully retaining the texture details of the image. In MRDDANet, shallow features are extracted from the noisy images by multiscale modules with different kernel sizes, and then, the extracted shallow features are mapped to the residual dense dual-attention network to obtain the deep features of SAR image. Finally, the final denoising image is generated through global residual learning. MRDDANet has advantages of both multiscale blocks and residual dense dual attention networks. The dense connection can fully extract features in the image, and the dual-channel attention enables MRDDANet to pay more attention to noise information, which is beneficial to remove noise and keep the details of the original image at the same time. Compared with state-of-the-art algorithms, the results of the experiment indicate that our method not only improves various objective indicators but also shows great advantages in visual effects.

Compared with traditional remoting image, there is a large amount of spectral information in the hyperspectral image (HSI), which makes HSI better reflect the actual condition of surface features. However, due to the limitations of imaging conditions, HSI tends to have a lower spatial resolution. In order to overcome this issue, we propose a spectral-spatial attention-based U-Net named SSAU-Net for HSI and multispectral image (MSI) fusion. The SSAU-Net constructs a spectral-spatial attention module by a coordinate-attention (CA) module and an efficient pyramid split attention (ESPA) module, which can enhance the image’s spectral information and spatial information. Meanwhile, the proposed network fully extracts the shallow and deep features of the images, and finally generates high-resolution (HR) hyperspectral images. Compared with state-of-the-art HSI-MSI fusion methods, the experimental results verify that the proposed method has a better subjective and objective fusion effect.

Since electroencephalogram (EEG) signals can truly reflect human emotional state, emotion recognition based on EEG has turned into a critical branch in the field of artificial intelligence. Aiming at the disparity of EEG signals in various emotional states, we propose a new deep learning model named three-dimension convolution attention neural network (3DCANN) for EEG emotion recognition in this paper. The 3DCANN model is composed of spatio-temporal feature extraction module and EEG channel attention weight learning module, which can extract the dynamic relation well among multi-channel EEG signals and the internal spatial relation of multi-channel EEG signals during continuous period time. In this model, the spatio-temporal features are fused with the weights of dual attention learning, and the fused features are input into the softmax classifier for emotion classification. In addition, we utilize SJTU Emotion EEG Dataset (SEED) to appraise the feasibility and effectiveness of the proposed algorithm. Finally, experimental results display that the 3DCANN method has superior performance over the state-of-the-art models in EEG emotion recognition.

The automatic and accurate segmentation of the prostate cancer from the multi-modal magnetic resonance images is of prime importance for the disease assessment and follow-up treatment plan. However, how to use the multi-modal image features more efficiently is still a challenging problem in the field of medical image segmentation. In this paper, we develop a cross-modal self-attention distillation network by fully exploiting the encoded information of the intermediate layers from different modalities, and the generated attention maps of different modalities enable the model to transfer significant and discriminative information that contains more details. Moreover, a novel spatial correlated feature fusion module is further employed for learning more complementary correlation and non-linear information of different modality images. We evaluate our model in five-fold cross-validation on 358 MRI images with biopsy confirmed. Without bells and whistles, our proposed network achieves state-of-the-art performance on extensive experiments.

Object detection, as one of the most important environment perception tasks for traffic safety in intelligent transportation systems, has been widely investigated recently. However, most of the researches focus on the fully supervised scenario, and inevitably lead to model failure. With the continuous development of Zero-Shot Learning (ZSL) models, Generalized Zero-Shot Detection (GZSD) has attracted great attention due to its ability of detecting unseen objects. Many researchers tend to map the detected visual features to semantic attributes and then separate seen and unseen domains during inference. But they have ignore that the generative methods generally have higher performance than these visual-semantic mapping methods, and they have been confirmed from previous GZSL methods. In order to make up for the vacancy of GZSD in the generative methods, we propose an idea of using curriculum learning to generate more precise unseen visual features. And with the excellent performance of WGAN-based method in sample synthesis, we realize the function of using semantics to generate visual features for unseen domains. In addition, we also adopt part of the idea of meta-learning to progressively correct the capability of the generator for better mitigating domain shift problem during the generation process. Through the above ideas, we can detect both seen and unseen bounding boxes and classify them accurately, by combining with the excellent detection ability of Faster-RCNN. Extensive experimental results on two popular datasets, i.e., MSCOCO and KITTI, show that our proposed method can outperform the state-of-the-art methods.

With the rapid development of convolutional neural networks (CNNs), significant progress has been achieved in semantic segmentation. Despite the great success, such deep learning approaches require large scale real-world datasets with pixel-level annotations. However, considering that pixel-level labeling of semantics is extremely laborious, many researchers turn to utilize synthetic data with free annotations. But due to the clear domain gap, the segmentation model trained with the synthetic images tends to perform poorly on the real-world datasets. Unsupervised domain adaptation (UDA) for semantic segmentation recently gains an increasing research attention, which aims at alleviating the domain discrepancy. Existing methods in this scope either simply align features or the outputs across the source and target domains or have to deal with the complex image processing and post-processing problems. In this work, we propose a novel multi-level UDA model named Confidence-and-Refinement Adaptation Model (CRAM), which contains a confidence-aware entropy alignment (CEA) module and a style feature alignment (SFA) module. Through CEA, the adaptation is done locally via adversarial learning in the output space, making the segmentation model pay attention to the high-confident predictions. Furthermore, to enhance the model transfer in the shallow feature space, the SFA module is applied to minimize the appearance gap across domains. Experiments on two challenging UDA benchmarks “GTA5-to-Cityscapes” and “SYNTHIA-to-Cityscapes” demonstrate the effectiveness of CRAM. We achieve comparable performance with the existing state-of-the-art works with advantages in simplicity and convergence speed.

In this article, we focus on the dynamic facial emotion recognition from online video. We combine deep neural networks with transfer learning theory and propose a novel model named DT-EFER. In detail, DT-EFER uses GoogLeNet to extract the deep features of key images from video clips. Then to solve the dynamic facial emotion recognition scenario, the framework introduces transfer learning theory. Thus, to improve the recognition performance, model DT-EFER focuses on the differences between key images instead of those images themselves. Moreover, the time complexity of this model is not high, even if previous exemplars are introduced here. In contrast to other exemplar-based models, experiments based on two datasets, namely, BAUM-1s and Extended Cohn–Kanade, have shown the efficiency of the proposed DT-EFER model.

In this paper, a reliable online destination prediction methodology is presented. The destination prediction methodology consists of a novel sequential complete diameter distance limited clustering method and an ensemble of random forest classifiers employing a one-vs-rest binarization strategy. Through the use of a novel OvR Uncertainty metric, predictions with high uncertainty could be withheld, thus increasing the overall reliability of the predictions made. The methodology was validated on 778 journeys from two real non-commuter vehicles based in the UK. These datasets allowed the methodology to be tested on real, yet challenging-to-predict journeys and irregular driver behavior. The sequential complete diameter distance limited clustering method was found to be a fast and effective method for sequentially clustering GPS coordinates into clusters that correspond to geographical locations. Prediction results showed that while only an overall mean prediction accuracy of 52% and 34% could be achieved on the two datasets, mean prediction accuracy could be significantly increased to over 90% and 73% respectively by only providing predictions with low uncertainty.

Abstract The competitive segmentation of fuzzy clustering is utilized in a greater manner to deal with the local spatial information of input medical images. Fuzzy clustering favours lesions and tumour identification through the segmentation process where less accuracy attainment and time complexity might be instigated for the identification of oddities. To rectify the above‐said problems, a novel methodology that encapsulates the combination of unsupervised neural network and fuzzy clustering processes, which effortlessly distinguishes the lesion and tumour region in MR brain images is developed through this study. The initial process of the proposed algorithm employs the histogram‐based feature extraction of the input images; whereof, a feature vector selection is made for the operation of self‐organizing map (SOM), which is a neural network functionary that progresses through the mapping process. Modification regarding the membership function of fuzzy entropy clustering (MFEC) is done based on the entropy value of the input image that results in quicker convergence. Finally, the updated objective function of MFEC algorithm augments the SOM result. It is found that the proposed SOM based MFEC algorithm is superior to other traditional segmentation algorithms, which have rendered the better visible understanding of the image. Further, the end‐results of the algorithm are verified through the evaluation of quantity metrics using ground truth of the brain MR images. The proposed SOM based MFEC algorithm precisely provides 82.26% of Jaccard value and 90.05% of Dice Overlap Index value, and these values prove better brain slices segmentation and provide enormous help to radiologists during patient diagnosis.

Abstract Facial expression recognition has been a long‐standing problem in the field of computer vision. This paper proposes a new simple scheme for effective recognition of facial expressions based on a hybrid feature descriptor and an improved classifier. Inspired by the success of stationary wavelet transform in many computer vision tasks, stationary wavelet transform is first employed on the pre‐processed face image. The pyramid of histograms of orientation gradient features is then computed from the low‐frequency stationary wavelet transform coefficients to capture more prominent details from facial images. The key idea of this hybrid feature descriptor is to exploit both spatial and frequency domain features which at the same time are robust against illumination and noise. The relevant features are subsequently determined using linear discriminant analysis. A new least squares support vector machine parameter tuning strategy is proposed using a contemporary optimisation technique called Jaya optimisation for classification of facial expressions. Experimental evaluations are performed on Japanese female facial expression and the Extended Cohn–Kanade (CK+) datasets, and the results based on 5‐fold stratified cross‐validation test confirm the superiority of the proposed method over state‐of‐the‐art approaches.

Multivariate time series (MTS) prediction aims at predicting future time series by extracting multiple forms of dependencies of past time series. Traditional prediction methods and deep learning-based prediction methods focus on extracting the dynamic relationships of certain aspects of MTS, especially the temporal characteristics, often neglecting the spatial and temporal dynamic correlations of MTS. Inspired by convolution neural network (CNN) and attention mechanism, this paper proposes a convolution LSTM network model based on MTS prediction with two-stage attention. Specifically, we first propose a new MTS preprocessing method to perform convolution operations better. Then convolution layer extracts spatial correlation of MTS and LSTM model extracts temporal correlation. It is worth mentioning that the combination of attention mechanism and LSTM can effectively solve the problem of insufficient time dependency in MTS prediction. In addition, dual-stage attention mechanism can effectively eliminate irrelevant information, select the relevant exogenous sequence, give it higher weight, and increase the past value of the target sequence to further eliminate irrelevant information. Finally, the MTS spatio-temporal correlation is extracted to improve the prediction accuracy, and the model is interpreted. Experimental results show that the model has broad application prospects. Experiments based on typical datasets of finance, environment, and energy determine the optimal window size and hidden size of the prediction, and demonstrate that the model achieves the state-of-the-art effect compared to the other four deep learning models. On top of that, the model is not only suitable for single-step prediction of MTS, but also suitable for multistep prediction of time step in a certain range.

The seven papers in this special section focus on advanced machine learning methods for brain machine interfacing. Particular emphasis is on novel theories and methods using transfer learning and deep learning proposed for Brain-Machine Interfacing (BMI) or Brain-Computer Interfacing (BCI). Our purpose is to review the new progress and achievements on transfer learning, deep learning, and their applications in BMI or BCI in recent years.

(Aim) Breast cancer is the most common cancer in women and the second most common cancer worldwide. With the rapid advancement of deep learning, the early stages of breast cancer development can be accurately detected by radiologists with the help of artificial intelligence systems. (Method) Based on mammographic imaging, a mainstream clinical breast screening technique, we present a diagnostic system for accurate classification of breast abnormalities based on ResNet-50. To improve the proposed model, we created a new data augmentation framework called SCDA (Scaling and Contrast limited adaptive histogram equalization Data Augmentation). In its procedure, we first conduct the scaling operation to the original training set, followed by applying contrast limited adaptive histogram equalisation (CLAHE) to the scaled training set. By stacking the training set after SCDA with the original training set, we formed a new training set. The network trained by the augmented training set, was coined as ResNet-SCDA-50. Our system, which aims at a binary classification on mammographic images acquired from INbreast and MINI-MIAS, classifies masses, microcalcification as "abnormal", while normal regions are classified as "normal". (Results) We present the first attempt to use the image contrast enhancement method as the data augmentation method, resulting in an averaged 98.55 percent specificity and 92.83 percent sensitivity, which gives our best model an overall accuracy of 95.74 percent. (Conclusion) Our proposed method is effective in classifying breast abnormality.

Mixing patterns (MPs) in social trust networks (STNs) are increasingly attracting attention because they can assist analysts in designing information dissemination tactics and planning electronic word-of-mouth (eWOM) campaigns. However, the existing studies on MPs do not explain the assortative or disassortative tendencies of STNs due to their omission of the support of the sociological theory, as well as that of network theory. To address this issue, this study investigates the MPs in STNs from the standpoint of social identity theory (SIT). The user trust networks (UTNs) are modeled by a directed multigraph (DMG). Then, the structural properties of homogeneous trust networks and heterogeneous trust networks are explored via measures that include degree centrality, the correlation coefficient (CC), the cumulative distribution of the ratio of trust degree to distrust degree (CDRTD), and the assortativity coefficient. The MPs of homogeneous trust networks and heterogeneous trust networks are explained from the perspective of SIT. An experiential evaluation is conducted in the constructed homogeneous trust networks and heterogeneous trust networks using a real-world data set crawled from Epinions. The research findings indicate that the MPs in homogeneous trust networks tend toward assortative mixing (AM), and those in heterogeneous trust networks tend toward disassortative mixing (DM). The experimental results show that the performance of the proposed approach is superior to that of the state-of-the-art approach to influential user identification.

Inaccuracies in anomaly prediction have become an alarming issue in the field of medical image analysis, and these quandaries have burgeoned due to the errors caused by the operator, instrument/device and environment, whereof, these troubles can be unanimously rectified with the advent of a novel segmentation approach proposed through this article. The novel approach encapsulates the functionary of spatially constrained fish school optimization (SCFSO) algorithm and interval type-II fuzzy logic system (IT2FLS) techniques, which resolves the erroneous prediction of anomalies present in various topographical locations in brain subjects of magnetic resonance imaging (MRI) modality. Huge datasets and complex tumor (anomalies) can be intervened and examined with ease by the developed approach, and this could be a proactive measure for being implemented or incorporated in clinical practice for the betterment of both doctors and patients, and it can perpetuate a profound experience to the doctors. The developed SCFSO-IT2FLS technique was applied to BRATS-SICAS dataset, and the evaluation metrics—namely, the Dice overlap index and sensitivity value were delivered as 96 ± 2.1 and 98 ± 1.1 by the proposed technique. These values are better than the conventional techniques and the proposed technique is applicable for the segmentation of T1-weighted, T2-weighted, and fluid attenuated inversion recovery MRI sequences of various axes coordination. A clear extraction of tumor region from nontumor region (edema) is rendered by the proposed technique and a therapeutic preplanning could always be made with such a provision/advantage.

Traditional deep learning methods are sub-optimal in classifying ambiguity features, which often arise in noisy and hard to predict categories, especially, to distinguish semantic scoring. Semantic scoring, depending on semantic logic to implement evaluation, inevitably contains fuzzy description and misses some concepts, for example, the ambiguous relationship between normal and probably normal always presents unclear boundaries (normal - more likely normal - probably normal). Thus, human error is common when annotating images. Differing from existing methods that focus on modifying kernel structure of neural networks, this study proposes a dominant fuzzy fully connected layer (FFCL) for Breast Imaging Reporting and Data System (BI-RADS) scoring and validates the universality of this proposed structure. This proposed model aims to develop complementary properties of scoring for semantic paradigms, while constructing fuzzy rules based on analyzing human thought patterns, and to particularly reduce the influence of semantic conglutination. Specifically, this semantic-sensitive defuzzier layer projects features occupied by relative categories into semantic space, and a fuzzy decoder modifies probabilities of the last output layer referring to the global trend. Moreover, the ambiguous semantic space between two relative categories shrinks during the learning phases, as the positive and negative growth trends of one category appearing among its relatives were considered. We first used the Euclidean Distance (ED) to zoom in the distance between the real scores and the predicted scores, and then employed two sample t test method to evidence the advantage of the FFCL architecture. Extensive experimental results performed on the CBIS-DDSM dataset show that our FFCL structure can achieve superior performances for both triple and multiclass classification in BI-RADS scoring, outperforming the state-of-the-art methods.

Synthetic aperture radar (SAR) images often interfere with speckle artifacts that have a great impact on subsequent processing and analysis operations. To remove speckle artifacts, this article introduces a hybrid denoising approach by using a convolutional neural network (CNN) and consistent cycle spinning (CCS) in the nonsubsample shearlet transform (NSST) domain. First, we apply NSST to a noisy SAR image to gain low- and high-frequency coefficients. Second, we adopt a learned deep CNN model to eliminate the speckle noise in the low-frequency coefficients, which retains more contour information. Third, we employ CCS to enhance the high-frequency coefficients, which preserves more details of the original SAR image. Finally, we obtain the denoised image by using inverse NSST applied to the denoised coefficients. Compared with state-of-the-art algorithms, the results of the experiment indicate that our method not only achieves better speckle removal performance but also maintains more detailed information retention.

In this article, a new interpretation of the polar format algorithm (PFA) for general bistatic spotlight synthetic aperture radar (SAR) imaging is presented. From the viewpoint of 2-D decoupling, we examine the commonly adopted implementation of the PFA, i.e., the separable 1-D range and azimuth resampling procedures for their roles in range cell migration (RCM) correction, respectively. Utilizing this new formulation, we analyze the effect of range and azimuth resampling on the residual 2-D phase error and reveal the inherent structure characteristics of the residual 2-D phase error in the wavenumber domain. By exploiting the available a priori knowledge on the phase error structure, a structure-aided 2-D autofocus approach to refocus the defocused PFA imagery is proposed. The proposed approach fully exploits the potentiality of the available data and the a priori knowledge about the phase error that need to estimate, so the accuracy of the residual 2-D phase error estimation and correction can be greatly improved. Finally, experimental results are presented to show the effectiveness of the proposed approach.

The aim of this article is to develop a methodology for measuring thedegree of unpredictabilityin dynamical systems with memory, i.e., systems with responses dependent on a history of past states. The proposed model is generic, and can be employed in a variety of settings, although its applicability here is examined in the particular context of an industrial environment: gas turbine engines. The given approach consists in approximating the probability distribution of the outputs of a system with a deep recurrent neural network; such networks are capable of exploiting the memory in the system for enhanced forecasting capability. Once the probability distribution is retrieved, theentropyormissing informationabout the underlying process is computed, which is interpreted as the uncertainty with respect to the system's behavior. Hence, the model identifies how far the system dynamics are from its typical response, in order to evaluate the system reliability and to predict system faults and/ornormal accidents. The validity of the model is verified with sensor data recorded from commissioning gas turbines, belonging to normal and faulty conditions.

With the development of 5G communication and transportation infrastructure, transportation systems face challenges to serve future smart cities regarding effective operation and cost optimization for electric vehicle networks. Thus, heterogeneous networking optimization approaches for these vehicles have been investigated, which have great potential in real-time communications, intelligent processing, reliable understanding, and efficient management. The guest editors have selected 16 articles for review in this special issue. A summary of these articles is outlined below.

Semantic hashing enables computation and memory-efficient image retrieval through learning similarity-preserving binary representations. Most existing hashing methods mainly focus on preserving the piecewise class information or pairwise correlations of samples into the learned binary codes while failing to capture the mutual triplet-level ordinal structure in similarity preservation. In this article, we propose a novel Probability Ordinal-preserving Semantic Hashing (POSH) framework, which for the first time defines the ordinal-preserving hashing concept under a non-parametric Bayesian theory. Specifically, we derive the whole learning framework of the ordinal similarity-preserving hashing based on the maximum posteriori estimation, where the probabilistic ordinal similarity preservation, probabilistic quantization function, and probabilistic semantic-preserving function are jointly considered into one unified learning framework. In particular, the proposed triplet-ordering correlation preservation scheme can effectively improve the interpretation of the learned hash codes under an economical anchor-induced asymmetric graph learning model. Moreover, the sparsity-guided selective quantization function is designed to minimize the loss of space transformation, and the regressive semantic function is explored to promote the flexibility of the formulated semantics in hash code learning. The final joint learning objective is formulated to concurrently preserve the ordinal locality of original data and explore potentials of semantics for producing discriminative hash codes. Importantly, an efficient alternating optimization algorithm with the strictly proof convergence guarantee is developed to solve the resulting objective problem. Extensive experiments on several large-scale datasets validate the superiority of the proposed method against state-of-the-art hashing-based retrieval methods.

No abstract available.

Aim: This study proposes a new artificial intelligence model based on cardiovascular computed tomography for more efficient and precise recognition of Tetralogy of Fallot (TOF). Methods: Our model is a structurally optimized stochastic pooling convolutional neural network (SOSPCNN), which combines stochastic pooling, structural optimization, and convolutional neural network. In addition, multiple-way data augmentation is used to overcome overfitting. Grad-CAM is employed to provide explainability to the proposed SOSPCNN model. Meanwhile, both desktop and web apps are developed based on this SOSPCNN model. Results: The results on ten runs of 10-fold cross-validation show that our SOSPCNN model yields a sensitivity of 92.25±2.19, a specificity of 92.75±2.49, a precision of 92.79±2.29, an accuracy of 92.50±1.18, an F1 score of 92.48±1.17, an MCC of 85.06±2.38, an FMI of 92.50±1.17, and an AUC of 0.9587. Conclusion: The SOSPCNN method performed better than three state-of-the-art TOF recognition approaches.

Crowd counting is of considerable significance to society in terms of public safety and urban development. Manual counting of people in a video or photo is often time-consuming and labour-intensive. People will need an efficient and economy way instead of counting manually. Nowadays, the convolutional neural network was popularly utilized as the baseline for crowd counting. However, the more complex the CNN-based algorithm, the more computing resources will be consumed. This article aims to present a simpler and faster fully optimized convolutional neural network for crowd counting with desired performance. To minimize the computational cost on training networks, we proposed a fully optimized method to build our network. Extensive experiments on our fully optimized convolutional neural network indicate the superiority of our network that has very high accuracy and speed on small scale crowd.

Summary Lung cancer has become one of the major diseases that seriously threaten human health. Early diagnosis of lung cancer is very important for the quality of lung cancer patients. Researching computer‐assisted systems for lung cancer recognition is one of the effective means to help physicians quickly diagnose lung cancer. Therefore, this paper studies the multimodal recognition algorithm for lung images. This algorithm uses CT images, PET images, and PET/CT images of lung tumors as the experimental objects and uses convolutional neural network to realize lung tumor recognition. Based on the recognition of lung tumors by convolutional neural networks, an integrated convolutional neural network was used to identify lung tumors to improve the recognition accuracy and reduce the training time. The experimental results show that the convolutional neural network can effectively identify the lung tumor images. The number of iterations and batch size in the training process will have an impact on the recognition of lung tumors. The integrated convolutional neural network designed is superior to single convolutional neural network in terms of recognition accuracy and time consumption.

Summary Cerebral micro‐bleedings are small chronic brain hemorrhages caused by structural abnormalities of the small vessels. CMBs can be found from individuals with stroke at memory clinics and even healthy elderly people. CMBs indicate hemorrhage‐prone pathological states. Research shows that CMBs are associated with an increased risk of future ischemic stroke, intra‐cerebral hemorrhage (ICH), dementia, and death. Considering that CMBs severely influence people's life, it is necessary to identify the CMBs in an early stage to prevent from further deterioration and to help people live a healthy life. In this paper, we proposed using CNN with stochastic pooling for the CMB detection. CNN has good performance in image and video recognition, recommender system, and nature language processing. Based on the collected subject, the experiment result shows that the six‐convolution layer and three fully‐connected layer CNN, nine‐layers in total, achieved sensitivity, specificity, accuracy, and precision as 97.22%, and 97.35%, 97.28%, and 97.35% in average of ten runs, which shows better performance than five state‐of‐the‐art methods.

Segmentation of handwritten words into isolated characters and their recognition are challenging due to the presence of high variability and cursiveness in Indian scripts. The complex shapes and availability of numerous atomic character classes, compound characters, modifiers, ascendants, and descendants make the recognition task even more difficult. A holistic approach effectively tackles such issues by avoiding the character‐level segmentation and the earlier holistic methods have been mostly developed using multi‐stage machine learning architecture. In this study, a deep convolutional neural network‐based holistic method termed ‘H‐WordNet’ is proposed for handwritten word recognition. The H‐WordNet model includes merely four convolutional layers and one fully connected layer to effectively classify the word images', which lead to a significant reduction in parameters. The efficacy of different pooling operations with the proposed model is investigated. The main purpose of this study is to avoid the need for handcrafted feature extraction and obtain a more stable and generalised system for word recognition. The proposed model is evaluated using a standard handwritten Bangla word database (CMATERdb2.1.2), which contains 18000 Bangla word images of 120 different categories and it obtained a higher recognition accuracy of 96.17% when compared to recent state‐of‐the‐art methods.

This research study is intended to deliver effective magnetic resonance (MR) brain image segmentation, which is an ambiguous process in the domain of medical image analysis. In general, MR brain image comprises various tissue structures; and an accurate representation of the above‐mentioned regions is essential to have a perfect identification of different grades of tumours, and obtaining effective demarcation of different areas in which the oedema portion is widespread. The accurate representation and identification of the abnormal regions in the MR images can be a vital tool for the radiologists and oncologists to proceed further with the treatment processes. This study aims in developing a novel automated approach that combines self‐organising map and interval type‐2 fuzzy logic clustering, providing ample knowledge to the clinicians in identifying the aberrant regions present in the patient brain. A non‐invasive analysis blended with quicker segmentation results are proffered by the proposed methodology and its functioning abilities have been assessed using comparison metrics such as mean‐squared error (MSE), peak signal‐to‐noise ratio (PSNR), processing time duration, and few other standard metrics. The proposed methodology has offered commendable MSE and PSNR values, which are 0.234778 and 54.847 dB, and it can be undeniably utilised for analysing the patient diseases.

Artificial intelligence and all its supporting tools, e.g. machine and deep learning in computational intelligence-based systems, are rebuilding our society (economy, education, life-style, etc.) and promising a new era for the social welfare state. In this paper we summarize recent advances in data science and artificial intelligence within the interplay between natural and artificial computation. A review of recent works published in the latter field and the state the art are summarized in a comprehensive and self-contained way to provide a baseline framework for the international community in artificial intelligence. Moreover, this paper aims to provide a complete analysis and some relevant discussions of the current trends and insights within several theoretical and application fields covered in the essay, from theoretical models in artificial intelligence and machine learning to the most prospective applications in robotics, neuroscience, brain computer interfaces, medicine and society, in general.

We propose a new method for generating synthetic CT images from modified Dixon (mDixon) MR data. The synthetic CT is used for attenuation correction (AC) when reconstructing PET data on abdomen and pelvis. While MR does not intrinsically contain any information about photon attenuation, AC is needed in PET/MR systems in order to be quantitatively accurate and to meet qualification standards required for use in many multi-center trials. Existing MR-based synthetic CT generation methods either use advanced MR sequences that have long acquisition time and limited clinical availability or use matching of the MR images from a newly scanned subject to images in a library of MR-CT pairs which has difficulty in accounting for the diversity of human anatomy especially in patients that have pathologies. To address these deficiencies, we present a five-phase interlinked method that uses mDixon MR acquisition and advanced machine learning methods for synthetic CT generation. Both transfer fuzzy clustering and active learning-based classification (TFC-ALC) are used. The significance of our efforts is fourfold: 1) TFC-ALC is capable of better synthetic CT generation than methods currently in use on the challenging abdomen using only common Dixon-based scanning. 2) TFC partitions MR voxels initially into the four groups regarding fat, bone, air, and soft tissue via transfer learning; ALC can learn insightful classifiers, using as few but informative labeled examples as possible to precisely distinguish bone, air, and soft tissue. Combining them, the TFC-ALC method successfully overcomes the inherent imperfection and potential uncertainty regarding the co-registration between CT and MR images. 3) Compared with existing methods, TFC-ALC features not only preferable synthetic CT generation but also improved parameter robustness, which facilitates its clinical practicability. Applying the proposed approach on mDixon-MR data from ten subjects, the average score of the mean absolute prediction deviation (MAPD) was 89.78±8.76 which is significantly better than the 133.17±9.67 obtained using the all-water (AW) method (p=4.11E-9) and the 104.97±10.03 obtained using the four-cluster-partitioning (FCP, i.e., external-air, internal-air, fat, and soft tissue) method (p=0.002). 4) Experiments in the PET SUV errors of these approaches show that TFC-ALC achieves the highest SUV accuracy and can generally reduce the SUV errors to 5% or less. These experimental results distinctively demonstrate the effectiveness of our proposed TFCALC method for the synthetic CT generation on abdomen and pelvis using only the commonly-available Dixon pulse sequence.

No abstract available.

( Aim ) Multiple sclerosis is a neurological condition that may cause neurologic disability. Convolutional neural network can achieve good results, but tuning hyperparameters of CNN needs expert knowledge and are difficult and time-consuming. To identify multiple sclerosis more accurately, this article proposed a new transfer-learning-based approach. ( Method ) DenseNet-121, DenseNet-169, and DenseNet-201 neural networks were compared. In addition, we proposed the use of a composite learning factor (CLF) that assigns different learning factor to three types of layers: early frozen layers, middle layers, and late replaced layers. How to allocate layers into those three layers remains a problem. Hence, four transfer learning settings (viz., Settings A, B, C, and D) were tested and compared. A precomputation method was utilized to reduce the storage burden and accelerate the program. ( Results ) We observed that DenseNet-201-D (the layers from CP to T3 are frozen, the layers of D4 are updated with learning factor of 1, and the final new layers of FCL are randomly initialized with learning factor of 10) can achieve the best performance. The sensitivity, specificity, and accuracy of DenseNet-201-D was 98.27± 0.58, 98.35± 0.69, and 98.31± 0.53, respectively. ( Conclusion ) Our method gives better performances than state-of-the-art approaches. Furthermore, this composite learning rate gives superior results to traditional simple learning factor (SLF) strategy.

In this paper, a methodology for assessing the unpredictability of systems with memory was developed. The proposed approach consists in approximating the probability distribution exhibited by the response of a system, understood as a stochastic process, with a deep recurrent neural network; such networks offer increased forecasting capability by exploiting an accumulative register of previous system states. Once the probability distribution is computed, the uncertainty or entropy of the underlying process is measured. This measure determines the degree of regularity in the system, and identifies how atypical the system dynamics are. The proposed model was validated by identifying industrial gas turbine engine faults from recorded sensor data.

Due to recent computer technology advancements the artificial neural networks (ANN), essentially with the same architecture as 20 years ago, have changed their status within digital media area.Then, for instance the linear discriminant analysis, i.e.LDA based face recognition systems outperformed their ANN counterparts.In general ANN classifiers were comparable in efficiency with specialized to the particular problem solutions only if good features were "manually" designed as input of ANN.Now, the convolutional neural networks (CNN) automatically design the features on the basis of raw digital media objects delivered as tensors and processed further in the consecutive layers as tensors, as well.A deep cascade of convolutional layers (DNN) creates an application oriented feature extractor, operating only with small kernels, and followed by primitive nonlinearities such as rectified linear units (ReLU), or pooling filters.In digital media research, the current power of DNN "floods new islands" of applications which were reserved for specialized approaches.

The filtered backprojection (FBP) algorithm is a popular choice for complicated trajectory synthetic aperture radar (SAR) image formation processing due to its inherent nonlinear motion compensation capability. However, how to efficiently refocus the defocused FBP imagery when the motion measurement is not accurate enough is still a challenging problem. In this paper, a new interpretation of the FBP derivation is presented from the Fourier transform point of view. Based on this new viewpoint, the property of the residual 2-D phase error in FBP imagery is analyzed in detail. Then, by incorporating the derived a priori knowledge on the 2-D phase error, an accurate and efficient 2-D autofocus approach is proposed. This new approach performs the parameter estimation in a dimension-reduced parameter subspace by exploiting the a priori analytical structure of the 2-D phase error, therefore it possesses much higher accuracy and efficiency than the conventional blind methods. Finally, experimental results clearly demonstrate the effectiveness and robustness of the proposed method.

Noise detection accuracy is crucial in suppressing random-valued impulse noise. Both false and miss detections determine the final estimation performance. Deterministic detection methods, which distinctly classify pixels into noisy or uncorrupted pixels, tend to increase the estimation error because some uncorrupted edge points are hard to discriminate from the random-valued impulse noise points. This paper proposes an iterative structure-adaptive fuzzy estimation (SAFE) for random-valued impulse noise suppression. This SAFE method is developed in the framework of Gaussian maximum likelihood estimation. The structure-adaptive fuzziness is reflected by two structure-adaptive metrics based on pixel reliability and patch similarity, respectively. The reliability metric for each pixel (as noise free) is estimated via a novel-minimal-path-based structure propagation to give full consideration of the spatially varying image structures. A robust iteration stopping strategy is also proposed by evaluating the reestimation error of the uncorrupted intensity information. The comparative experimental results show that the proposed structure-adaptive fuzziness can lead to effective restoration. An efficient implementation of this SAFE method is also realized via graphics-processing-unit-based parallelization.

(Aim) In order to detect pathological brains in a more efficient way, (Method) we proposed a novel system of pathological brain detection (PBD) that combined wavelet packet Tsallis entropy (WPTE), feedforward neural network (FNN), and real-coded biogeography-based optimization (RCBBO). (Results) Th e experiments showed the proposed WPTE + FNN + RCBBO approach yielded an average accuracy of 99.49% over a 255-image dataset. (Conclusions) The WPTE + FNN + RCBBO performed better than 10 state-of-the-art approaches.

(Aim) Abnormal breast can be diagnosed using the digital mammography. Traditional manual interpretation method cannot yield high accuracy. (Method) In this study, we proposed a novel computer-aided diagnosis system for detecting abnormal breasts in mammogram images. First, we segmented the region-o f-interest. Next, the weighted-type fractional Fourier transform (WFRFT) was employed to obtain the unified time-frequency spectrum. Third, principal component analysis (PCA) was introduced and used to reduce the spectrum to only 18 principal components. Fourth, feed-forward neural network (FNN) was utilized to generate the classifier. Finally, a novel algorithm-specific parameter free approach, Jaya, was employed to train the classifier. (Results) Our proposed WFRFT + PCA + Jaya-FNN achieved sensitivity of 92.26% ± 3.44%, specificity of 92.28% ± 3.58%, and accuracy of 92.27% ± 3.49%. (Conclusions) The proposed CAD system is effective in detecting abnormal breasts and performs better than 5 state-of-the-art systems. Besides, Jaya is more effective in training FNN than BP, MBP, GA, SA, and PSO.

To detect the sensorineural hearing loss (SNHL) from healthy people accurately, we used magnetic resonance imaging (MRI) to obtain the imaging data, and then proposed a new computer-aided diagnosis (CAD) system, on the basis of texture analysis method. In the first, we extracted 12-element feature from each brain image via fractional Fourier entropy (FRFE). Afterwards, multilayer perceptron (MLP) was employed as the classifier, which was trained by a novel fitness-scaling adaptive genetic algorithm (FSAGA). The statistical analysis over 49 subjects showed the overall accuracy of our method yielded 95.51%. Experimental results performed better than four state-of-the-art weight optimization methods, and this CAD system give significantly better performance than manual interpretation.

We investigate local fractional nonlinear Riccati differential equations (LFNRDE) by transforming them into local fractional linear ordinary differential equations. The case of LFNRDE with constant coefficients is considered and non-differentiable solutions for special cases obtained.

The Fractional Fourier transform (FRFT) is a relatively novel linear transforms that is a generalization of conventional Fourier transform (FT). FRFT can transform a particular signal to a unified time-frequency domain. In this survey, we try to present a comprehensive investigation of FRFT. Firstl y, we provided definition of FRFT and its three discrete versions (weighted-type, sampling-type, and eigendecomposition-type). Secondly, we offered a comprehensive theoretical research and technological studies that consisted of hardware implementation, software implementation, and optimal order selection. Thirdly, we presented a survey on applications of FRFT to following fields: communication, encryption, optimal engineering, radiology, remote sensing, fractional calculus, fractional wavelet transform, pseudo-differential operator, pattern recognition, and image processing. It is hoped that this survey would be beneficial for the researchers studying on FRFT.

A non-differentiable resistor-capacitor circuit comprised of the capacitor and resistor in the fractal-time domain is first proposed in this article. The solution behavior of the corresponding local fractional ordinary differential equation is presented for the Mittag-Leffler decay defined on Canto r sets. The obtained results reveal the sufficiency of the local fractional calculus in the analysis of the fractal electrical systems.

Background: We proposed a new automatic and rapid computer-aided diagnosis system to detect pathological brain images obtained in the scans of magnetic resonance imaging (MRI). Methods: For simplification, we transformed the problem to a binary classification task (pathological or normal). It consisted of two steps: first, Hu moment invariants (HMI) were extracted from a specific MR brain image; then, seven HMI features were fed into two classifiers: twin support vector machine (TSVM) and generalised eigenvalue proximal SVM (GEPSVM). Results: Then, a 5 × 5-fold cross validation on a data set containing 90 MR brain images, demonstrated that the proposed methods “HMI + GEPSVM” and “HMI + TSVM” achieved classification accuracy of 98.89%, higher than eight state-of-the-art methods: “DWT + PCA + BP-NN”, “DWT + PCA + RBF-NN”, “DWT + PCA + PSO-KSVM”, “WE + BP-NN”, “WE + KSVM”, “DWT + PCA + GA-KSVM”, “WE + PSO-KSVM” and “WE + BBO-KSVM”. Conclusion: The proposed methods are superior to other methods on pathological brain detection (p < 0.05).

In recent years, sonar has been widely used for restoring the underwater terrain. Sonar imaging has the benefits such as long-range photographing, robust for turbidity water. However, it is not suitable for short-range imaging. Meanwhile, it also cannot meet the need of mining machine. Therefore, it is important to develop a 3D reconstruction method for short-range imaging. In this paper, we propose a Kinect-based underwater 3D image reconstruction method. To overcome the drawbacks of low accuracy of depth maps, we propose a novel super-resolution (SR) method, which uses the underwater dark channel prior dehazing, weight guided image SR, and inpainting. The proposed method considered the influence of mud sediments in water, it performs better than the traditional methods. The experimental results demonstrated that, after inpainting, dehazing and the super-resolution, it can obtain high accuracy depth maps.

In order to detect cerebral microbleed more efficiently, we developed a novel computer-aided detection method based on susceptibility-weighted imaging. We enrolled five CADASIL patients and five healthy controls. We used a 20x20 neighboring window to generate samples on each slice of the volumetric brain images. The sparse autoencoder (SAE) was used to unsupervised feature learning. Then, a deep neural network was established using the learned features. The results over 10x10-fold cross validation showed our method yielded a sensitivity of 93.20±1.37%, a specificity of 93.25±1.38%, and an accuracy of 93.22±1.37%. Our result is better than Roy's method, which was proposed in 2015.

Abstract Accurate fruit classification is difficult to accomplish because of the similarities among the various categories. In this paper, we proposed a novel fruit‐classification system, with the goal of recognizing fruits in a more efficient way. Our methodology included the following steps. First, a four‐step pre‐processing was employed. Second, the features (colour, shape, and texture) were extracted. Third, we utilized principal component analysis to remove excessive features. Fourth, a novel fruit‐classification system based on biogeography‐based optimization (BBO) and feedforward neural network (FNN) was proposed, with the short name of BBO‐FNN. The experiment employed over 1653 chromatic fruit images (18 categories) by fivefold stratified cross‐validation. The results showed that the proposed BBO‐FNN yielded an overall accuracy of 89.11%, which was higher than the five state‐of‐the‐art methods: genetic algorithm‐FNN, artificial bee colony‐FNN, particle swarm optimization‐FNN, kernel support vector machine, and ant colony optimization‐FNN. Also, the BBO‐FNN achieved the same accuracy as fitness‐scaling chaotic artificial bee colony‐FNN, but it performed much faster than the latter. The proposed BBO‐FNN was effective in fruit‐classification in terms of classification accuracy and computation time. This indicated that it can be applied in credible use.

Minimal path techniques can efficiently extract geometrically curve-like structures by finding the path with minimal accumulated cost between two given endpoints. Though having found wide practical applications (e.g., line identification, crack detection, and vascular centerline extraction), minimal path techniques suffer from some notable problems. The first one is that they require setting two endpoints for each line to be extracted (endpoint problem). The second one is that the connection might fail when the geodesic distance between the two points is much shorter than the desirable minimal path (shortcut problem). In addition, when connecting two distant points, the minimal path connection might become inefficient as the accumulated cost increases over the propagation and results in leakage into some non-feature regions near the starting point (accumulation problem). To address these problems, this paper proposes an approach termed minimal path propagation with backtracking. We found that the information in the process of backtracking from reached points can be well utilized to overcome the above problems and improve the extraction performance. The whole algorithm is robust to parameter setting and allows a coarse setting of the starting point. Extensive experiments with both simulated and realistic data are performed to validate the performance of the proposed method.

Locally linear embedding is often invalid for sparse data sets because locally linear embedding simply takes the reconstruction weights obtained from the data space as the weights of the embedding space. This paper proposes an improved local linear embedding for sparse data sets. In the proposed method, the neighborhood correlation matrix presenting the position information of the points constructed from the embedding space is added to the correlation matrix in the original space, thus the reconstruction weights can be adjusted. As the reconstruction weights adjusted gradually, the position information of sparse points can also be changed continually and the local geometry of the data manifolds in the embedding space can be well preserved. Experimental results on both synthetic and real-world data show that the proposed approach is very robust against sparse data sets.