Search

Gesture detection has recently attracted a lot of attention due to its wide range of applications, notably in human–computer interaction (HCI). However, when it comes to video-based gesture recognition, elements in the background unrelated to gestures slow down the system's classification rate. This paper presents an algorithm designed for the recognition of large-scale gestures. In the training phase, we utilize RGB-D videos, where the depth modality videos are derived from RGB modality videos using UNET and subsequently employed for testing. However, it's worth noting that in real-time applications of the proposed dynamic hand gesture recognition (DHGR) system, only RGB modality videos are needed. The algorithm begins by creating two dynamic images: one from the estimated depth video and the other from the RGB video. Dynamic images generated from RGB video excel in capturing spatial information; while, those derived from depth video excel in encoding temporal aspects. These two dynamic images are merged to form an RGB-D dynamic image (RDDI). The RDDI is then fed into a modified Xception-based CNN model for the purpose of gesture classification and recognition. In order to evaluate the system's performance, we conducted experiments using the EgoGesture and MSR Gesture datasets. The results are highly promising, with a reported classification accuracy of 91.64% for the EgoGesture dataset and an impressive 99.41% for the MSR Gesture dataset. The results demonstrated that the suggested system outperformed some existing techniques. [ABSTRACT FROM AUTHOR]

Power image tampering brings certain security risks to the safe operation of power grids, among which splicing tampering is the most common. Although image tampering detection has received much attention in recent years, relatively little research has been conducted on the practical application to power systems, and the detection results are poor due to the difficulty of learning subtle edge features of tampered regions and the lack of power image tampering dataset for training. In order to effectively detect image tampering regions, this paper proposes a splicing tampering detection model with a Dual-channel Augmented Attentive Dense-convolutional Network (DAAD-Net) structure. The model consists of three main parts: backbone network feature extraction, augmented attention feature extraction, and tampered region detection. Firstly, the backbone network feature extraction module fuses the original tampered image features with the image residual features and transfers them into the backbone network to extract the feature map. Secondly, the augmented attention feature extraction module extracts the tampered region features in the higher and lower layers using hierarchical encoding and decoding operations. Finally, the extracted feature maps of each layer are sent to the tampered region detection module, and by combining the loss of each feature map for optimizing the network parameters. Additionally, we produced a power image tampering dataset containing 552 samples. Experiments demonstrate that the proposed method outperforms current state-of-the-art models, with 1% to 31% improvement in evaluation indicators, and has good robustness to noise and JPEG compression attacks. [ABSTRACT FROM AUTHOR]

Breast cancer, a widespread global disease, represents a significant threat to women's health and lives. Many researchers have proposed computer-aided diagnosis systems for classifying breast cancer. The majority of the approaches primarily utilize deep learning methods, which overlook the crucial necessity of incorporating both local information for precise tumor detection. In addition, available breast cancer datasets are imbalanced in nature. Therefore, this paper presents the hybrid breast network (HBNet) for detection of breast cancer, designed to address two critical challenges: class imbalance and incorporation of both global and local information in achieving precise tumor classification. To overcome the problem of class imbalance, HBNet incorporates the borderline synthetic minority oversampling technique. Simultaneously, it employs a feature fusion approach to combine deep and handcrafted features extracted by utilizing ResNet50 and HOG which incorporates global and local information. Moreover, the proposed method integrates the block-matching and 3D denoising filter to effectively eliminate multiplicative noise that has enhanced the performance of the system. The proposed HBNet is evaluated with BUSI and UDIAT datasets and achieved an average accuracy of 95.824% and 90.37%, respectively. [ABSTRACT FROM AUTHOR]

Microscopic emission estimation based on driving states plays a crucial role in controlling the pollution of on-road vehicles. Existing research has evolved from fitting nonlinear models of driving cycles and emission factors to utilizing neural networks to exploit driving patterns and construct a nonlinear mapping between driving states and emission values. However, due to the small percentage of driving-cycle related to high emissions, it is still challenging to capture the vehicle emission peaks, which lead to the most noteworthy high-emission characteristics being regarded as abnormal disturbances instead. To address the issue, this paper proposes a peak-sensitive microscopic emission estimation framework characterized by sequence-to-sequence learning for on-road vehicles. Sequence-to-sequence learning emphasizes serial pattern mapping from driving sequences to emission sequences to provide more statistical constraints and reduce the estimation uncertainty brought by unstable driving behaviors. Specifically, the framework aims at capturing context features related to high emissions from sequences dynamically in an adaptive and self-learning way and is composed of a driving states embedding module and a dynamic aggregation module. Particularly, an incremental tracking loss (ITL) is proposed to adjust the incremental emissions at adjacent time steps by supervising the differences of the generated sequences, enabling the model to track sudden changes in emissions. Extensive experiments are conducted on the on-board diagnostics (OBD) dataset with 12628 sampling records collected from a heavy-duty diesel vehicle. The results show that the estimation accuracy of our proposed method is significantly better than state-of-the-art methods, and it can effectively capture high-emission peaks. [ABSTRACT FROM AUTHOR]

Establishing a reliable credit card fraud detection model has become a primary focus for academia and the financial industry. The existing anti-fraud methods face challenges related to low recall rates, inaccurate results, and insufficient causal modeling ability. This paper proposes a credit card fraud detection model based on counterfactual data enhancement of the triplet network. Firstly, we convert the problem of generating optimal counterfactual explanations (CFs) into a policy optimization of agents in the discrete–continuous mixed action space, thereby ensuring the stable generation of optimal CFs. The triplet network then utilizes the feature similarity and label difference of positive example samples and CFs to enhance the learning of the causal relationship between features and labels. Experimental results demonstrate that the proposed method improves the accuracy and robustness of the credit card fraud detection model, outperforming existing methods. The research outcomes are of significant value for both credit card anti-fraud research and practice while providing a novel approach to causal modeling issues across other fields. [ABSTRACT FROM AUTHOR]

For market players and policy officials, commodity price forecasts are crucial problems that are challenging to address due to the complexity of price time series. Given its strategic importance, corn crops are hardly an exception. The current paper evaluates the forecasting issue for China's weekly wholesale price index for yellow corn from January 1, 2010 to January 10, 2020. We develop a Gaussian process regression model using cross validation and Bayesian optimizations over various kernels and basis functions that could effectively handle this sophisticated commodity price forecast problem. The model provides precise out-of-sample forecasts from January 4, 2019 to January 10, 2020, with a relative root mean square error, root mean square error, and mean absolute error of 1.245%, 1.605, and 0.936, respectively. The models developed here might be used by market players for market evaluations and decision-making as well as by policymakers for policy creation and execution. [ABSTRACT FROM AUTHOR]

In order to realize the intelligent recognition and statistics of hydrogenation catalyst image information, this paper presents a new method to judge the active phase by image recognition, which is different from traditional methods. Firstly, considering that hydrogenation catalyst image targets are small and easy to stack, the feature extraction network in the CenterNet model is optimized by adding the multi-feature fusion module to improve the accuracy of the network in edge positioning. Secondly, according to the linear shape of the hydrogenation catalyst, the mask branch is added to the CenterNet model to train the hydrogenation catalyst stripes with unclear target to reduce the leakage rate of the hydrogenation catalyst. The experimental results show that the detection accuracy of the improved CenterNet network is 91 % , 7 % higher than that of the original one, with a decline in detection rate by 12 % . The method proposed in this paper can accurately identify and segment the hydrogenation catalyst in the electron microscope image, which can provide technical support for the statistics and analysis of the hydrogenation catalyst image. [ABSTRACT FROM AUTHOR]

Due to the growing demand for clean and sustainable energy sources, there has been an increasing interest in solar cells and photovoltaic panels. Nevertheless, determining the right design parameters to achieve the most efficient energy output that aligns with the energy system's needs can be quite challenging. This complexity arises from the intricate models and the inherent inaccuracies in the available information. To tackle this challenge, this paper introduces the adaptive sine–cosine particle swarm optimization algorithm (ASCA-PSO) as a method for estimating the parameters of solar cells and photovoltaic modules. The ASCA-PSO approach combines the strengths of the SCA and PSO algorithms in a two-tier process. In this process, SCA search agents explore the search space, while the PSO search agents leverage the outcomes derived from SCA exploration. This study evaluates the effectiveness of ASCA-PSO in accurately estimating the parameters of single- and double-diode models using data from two commercial solar cells. The findings are compared with those of cutting-edge methods. It is demonstrated that ASCA-PSO can identify global solutions for multifaceted and intricate objective functions. Furthermore, it proves to be a viable option for designing solar cells even in the presence of noise. [ABSTRACT FROM AUTHOR]

Medical image segmentation is crucial in using digital images for disease diagnosis, particularly in post-processing tasks such as analysis and disease identification. Segmentation of magnetic resonance imaging (MRI) and computed tomography images pose distinctive challenges attributed to factors such as inadequate illumination during the image acquisition process. Multilevel thresholding is a widely adopted method for image segmentation due to its effectiveness and ease of implementation. However, the primary challenge lies in selecting the optimal set of thresholds to achieve accurate segmentation. While Otsu's between-class variance and Kapur's entropy assist in identifying optimal thresholds, their application to cases requiring more than two thresholds can be computationally intensive. Meta-heuristic algorithms are commonly employed in literature to calculate the threshold values; however, they have limitations such as a lack of precise convergence and a tendency to become stuck in local optimum solutions. In this paper, we introduce an improved chameleon swarm algorithm (ICSA) to address these limitations. ICSA is designed for image segmentation and global optimization tasks, aiming to improve the precision and efficiency of threshold selection in medical image segmentation. ICSA introduces the concept of the "best random mutation strategy" to enhance the search capabilities of the standard chameleon swarm algorithm (CSA). This strategy leverages three distribution functions—Levy, Gaussian, and Cauchy—for mutating search individuals. These diverse distributions contribute to improved solution quality and help prevent premature convergence. We conduct comprehensive experiments using the IEEE CEC'20 complex optimization benchmark test suite to evaluate ICSA's performance. Additionally, we employ ICSA in image segmentation, utilizing Otsu's approach and Kapur's entropy as fitness functions to determine optimal threshold values for a set of MRI images. Comparative analysis reveals that ICSA outperforms well-known metaheuristic algorithms when applied to the CEC'20 test suite and significantly improves image segmentation performance, proving its ability to avoid local optima and overcome the original algorithm's drawbacks. Medical image segmentation is essential for employing digital images for disease diagnosis, particularly for post-processing activities such as analysis and disease identification. Due to poor illumination and other acquisition-related difficulties, radiologists are especially concerned about the optimal segmentation of brain magnetic resonance imaging (MRI). Multilevel thresholding is the most widely used image segmentation method due to its efficacy and simplicity of implementation. The issue, however, is selecting the optimum set of criteria to effectively segment each image. Although methods like Otsu's between-class variance and Kapur's entropy help locate the optimal thresholds, using them for more than two thresholds requires a significant amount of processing resources. Meta-heuristic algorithms are commonly employed in literature to calculate the threshold values; however, they have limitations such as a lack of precise convergence and a tendency to become stuck in local optimum solutions. Due to the aforementioned, we present an improved chameleon swarm algorithm (ICSA) in this paper for image segmentation and global optimization tasks to be able to address these weaknesses. In the ICSA method, the best random mutation strategy has been introduced to improve the searchability of the standard CSA. The best random strategy utilizes three different types of distribution: Levy, Gaussian, and Cauchy to mutate the search individuals. These distributions have different functions, which help enhance the quality of the solutions and avoid premature convergence. Using the IEEE CEC'20 test suite as a recent complex optimization benchmark, a comprehensive set of experiments is carried out in order to evaluate the ICSA method and demonstrate the impact of combining the best random mutation strategy with the original CSA in improving both the performance of the solutions and the rate at which they converge. Furthermore, utilizing the Otsu approach and Kapur's entropy as a fitness function, ICSA is used as an image segmentation method to select the ideal threshold values for segmenting a set of MRI images. Within the experiments, the ICSA findings are compared with well-known metaheuristic algorithms. The comparative findings showed that ICSA performs better than other competitors in solving the CEC'20 test suite and has a significant performance boost in image segmentation. [ABSTRACT FROM AUTHOR]

Successful deployment of convolutional neural networks on resource-constrained hardware platforms is challenging for ubiquitous AI applications. For latency-sensitive scenarios, real-time inference requires model compression techniques such as network pruning to achieve the purpose of inference acceleration. However, many researches focus on hardware-independent filter pruning methods, which cannot balance the contribution of pruned structure to latency and the accuracy drop. Although some pruning methods have introduced latency constraints into the pruning process, most of them are based on look-up tables, which omits the key step of hardware optimization, resulting in significant deviation in latency estimation. In this paper, we propose a novel latency-constrained pruning method, named hardware-in-loop pruning (HILP). It is based on the fast optimal pruning rate search within the layer and layer-wise hybrid pruning, which can prioritize removing the less important layers with considerable latency contributions. The proposed hardware-in-loop pipeline enables the hardware optimization module to be integrated into the entire framework. During pruning, an intermediate network architecture is automatically transformed to a deployable model for accurate latency measurement. The latency-optimized intermediate architecture is then selected by traversing all layers for next progressive step. HILP is generally applicable to any platform that provides a hardware optimization toolchain, such as NVIDIA GPU and Cambricon NPU. We evaluate HILP on both image classification task using ResNet50 with ImageNet and object detection task using YOLOv3 with COCO, HILP can reduce the inference latency of these two networks to 60% and 75%, respectively, within the range of accuracy variation not exceeding 0.6%. Extensive experiment results have proven that HILP is able to achieve a significant advantage in latency-accuracy performance compared to state-of-the-art methods. [ABSTRACT FROM AUTHOR]

This paper proposes a novel incipient fault estimation and fault-tolerant control approach for wireless power transfer (WPT) systems with disturbances and incipient sensor faults. Firstly, the dynamic models of the WPT system and incipient faults are established. Then, a k-step incipient fault estimation observer method is analyzed for estimating the states and incipient faults of the system. Based on it, a nonlinear dynamic output feedback fault-tolerant controller is devised to ensure the stability of the WPT system when considering incipient faults and disturbances. Further, the designed controller can monitor the system in real time without the knowledge of the system states. Besides, the stability analysis approach is rigorous, and the proof method can also apply to other similar systems. At last, the simulation example shows that the proposed fault diagnosis method is correct and effective. [ABSTRACT FROM AUTHOR]

One of the most popular techniques for changing the purpose of an image or resizing a digital image with content awareness is the seam-carving method. The performance of image resizing algorithms based on seam machining shows that these algorithms are highly dependent on the extraction of importance map techniques and the detection of salient objects. So far, various algorithms have been proposed to extract the importance map. In this paper, a new method based on Rényi entropy is proposed to extract the importance map. Also, a deep learning network has been used to detect salient objects. The simulator results showed that combining Rényi's importance map with a deep network of salient object detection performed better than classical seam-carving and other extended seam-carving algorithms based on deep learning. [ABSTRACT FROM AUTHOR]

Deep neural networks (DNNs) have been applied in many pattern recognition or object detection applications. DNNs generally consist of millions or even billions of parameters. These demanding computational storage and requirements impede deployments of DNNs in resource-limited devices, such as mobile devices, micro-controllers. Simplification techniques such as pruning have commonly been used to slim DNN sizes. Pruning approaches generally quantify the importance of each component such as network weight. Weight values or weight gradients in training are commonly used as the importance metric. Small weights are pruned and large weights are kept. However, small weights are possible to be connected with significant weights which have impact to DNN outputs. DNN accuracy can be degraded significantly after the pruning process. This paper proposes a roulette wheel-like pruning algorithm, in order to simplify a trained DNN while keeping the DNN accuracy. The proposed algorithm generates a branch of pruned DNNs which are generated by a roulette wheel operator. Similar to the roulette wheel selection in genetic algorithms, small weights are more likely to be pruned but they can be kept; large weights are more likely to be kept but they can be pruned. The slimmest DNN with the best accuracy is selected from the branch. The performance of the proposed pruning algorithm is evaluated by two deterministic datasets and four non-deterministic datasets. Experimental results show that the proposed pruning algorithm generates simpler DNNs while DNN accuracy can be kept, compared to several existing pruning approaches. [ABSTRACT FROM AUTHOR]

Generalized zero-shot learning (GZSL) aims to identify the samples that belong to a different label space than training samples, while the samples of these unseen classes have the same semantic space as the samples of seen classes. Therefore, most GZSL models use the samples to train the fundamental part, then use semantic information to match or generate the features of the unseen classes as the generalized part. However, the importance of features and their consistency with semantics are different. This difference causes jeopardizing between each other, becoming one of the main reasons that the classification results of seen and unseen classes in the GZSL are lower than the traditional classification and the CZSL. To maintain the training effect from two parts without jeopardizing each other, IDN-GZSL is presented in this paper. We provide rearrangement modules in different isolated scenarios to distinguish the positive and negative parts of the feature and extract the positive features. Three rearrangement modules have been trained using existing training data and generated data in corresponding scenarios. Although the three scenarios are isolated, the decoupling and relation modules are used to support the rearrangement and acquire positive features for semantic information, and the distillation module is used to establish indirect connections. The results demonstrate that our methods take effect with three significant improvements in ablation experiments: successfully trained rearrangement modules for obtaining positive features in multiple scenarios for classification, eliminated jeopardizing among different scenarios by establishing three isolation scenarios, and established necessary connections among them by distillation module. Our method performs the best classification ability of unseen classes on four datasets and achieves the best performance on AWA2, CUB, and FLO when evaluating the method through harmonic means. [ABSTRACT FROM AUTHOR]

Detectors at the Laser Interferometer Gravitational Wave Observatory (LIGO), along with its sister detectors Virgo and Kamioka Gravitational Wave Detector (KAGRA), endlessly collect data to be analyzed in search of gravitational wave signals (GW signals) produced primarily by massive merger events involving colliding black holes and neutron stars. These detectors are the foundation upon which the new field of gravitational wave astronomy is built. However, the algorithms used to sift through this incoming data are extremely computationally expensive. They constantly run the risk of lagging behind the detectors, which would be catastrophic for astronomers searching for merger events, and their inherent rigidity becomes an obstruction to the expansion of these detectors' range and sensitivity. To simplify and expedite the process of detecting and isolating GW signals, this paper presents a neural network (NN) cascade to automatically isolate GW signals in raw LIGO data. Perfecting such a NN system would improve the speed and efficiency of detecting and analyzing GW signals. This study uses a 2-stage cascade of convolutional autoencoders (CAEs) to detect and reconstruct GW signals buried in LIGO data, finding that it is effective at accurately isolating GW signals with very low latency on just one graphics processing unit (GPU). Ultimately, it is found that it is practical and likely beneficial for astronomers to use such a cascade to process incoming LIGO data. [ABSTRACT FROM AUTHOR]

In this paper, an online event-triggered self-learning scheme based on adaptive dynamic programming (ADP) is developed to address tracking control design for nonlinear systems with constrained input and uncertain disturbance. Firstly, the value function with non-quadratic function is defined for the augmented nominal system, and the constrained robust tracking problem is equivalent to the optimal control for solving the tracking event-triggered Hamilton–Jacobi–Bellman (ETHJB) equation. Then, a single-critic network is developed to obtain the value function and control law related to the solution of the tracking ETHJB equation, greatly reducing approximation errors and computational costs. To alleviate the requirement for the entire state sampling, we propose a triggering rule that ensures system stability while limiting control updates. Theoretical proof demonstrates that the tracking state of the closed-loop system and the weight approximation error of the neural network are uniformly ultimately bounded (UUB). Finally, two examples are provided to validate the availability of the proposed scheme. [ABSTRACT FROM AUTHOR]

Recently, fully transformer-based trackers have achieved impressive tracking results, but this also brings a great deal of computational complexity. Some researchers have applied token pruning techniques to fully transformer-based trackers to diminish the computational complexity, but this leads to missing contextual information that is important for the regression task in the tracker. In response to the above issue, this paper proposes a token fusion method that speeds up inference while avoiding information loss and thus improving the robustness of the tracker. Specifically, the input of the transformer's encoder contains search tokens and exemplar tokens, and the search tokens are divided into tracking object tokens and background tokens according to the similarity between search tokens and exemplar tokens. The tokens with greater similarity to the exemplar tokens are identified as tracking object tokens, and those with smaller similarity to the exemplar tokens are identified as background tokens. The tracking object tokens contain the discriminative features of the tracking object, for the sake of making the tracker pay more attention to the tracking object tokens while reducing the computational effort. All the tracking object tokens are kept, and then, the background tokens are weighted and fused to form new background tokens according to the attention weight of the background tokens to prevent the loss of contextual information. The token fusion method presented in this paper not only provides efficient inference of the tracker but also makes the tracker more robust. Extensive experiments are carried out on popular tracking benchmark datasets to verify the validity of the token fusion method. [ABSTRACT FROM AUTHOR]

The online content, including live streaming and short videos, provides abundant visual and textual product information to users, which offers insights into users' multiple and changeable preferences toward product style, brand, color, etc. These dynamic preferences are helpful to the product recommender system to extract users' purchasing patterns. However, existing sequential recommendation fails to combine with items' multimodal characteristics and dynamic preferences. This paper proposes the Online Content-Based Sequence Recommendation method (named OCBSR), where dynamic preferences and multimodal representation such as visual images, textual descriptions, and item-ID of products are leveraged with convolutional autoencoder to map multimodal data into a unified latent space. We devise a multimodal transformer mechanism to aggregate the multimodal representation. A novel contrastive infomax loss is constructed through contrastive prototype representations and noise-contrastive sequences. The loss enhances the correlation between single-modality and multimodality preferences and learns the contrastive representation. Then, users' dynamic preferences are extracted from the contrastive representation in each time step using the gated recurrent unit with skip connections. Experiments are conducted on two real-life multimodal datasets, H&M and KuaiRec. The hit ratio (HR) and normalized discounted cumulative gain (NDCG) results indicate that our approach noticeably outperforms the current advanced recommenders. The HR increases by above 20% and 25% on H&M and KuaiRec datasets, respectively; the NDCG rises about 12% and 24% on H&M and KuaiRec datasets, respectively. The ablation study verifies each module's effectiveness. This research contributes to online content literature and multimodal sequential recommendation in practice. [ABSTRACT FROM AUTHOR]

In recent years, the proliferation of vehicles on roadways, accompanied by an escalating demand for augmented safety and refined traffic management, has fueled substantial progress in the domains of computer vision and deep learning methodologies. This technological surge has precipitated ground-breaking applications, with real-time vehicle detection emerging as a linchpin. Capability to accurately detect vehicles in a real-time holds profound implications across diverse domains. In the context of autonomous driving, it underpins essential functions such as path planning, collision avoidance, and decision-making, essential for the safe operation of autonomous vehicles. In the arena of traffic monitoring, it empowers transportation authorities to acquire real-time insights into traffic dynamics, congestion patterns, and incident detection, thereby optimizing traffic flow and resource allocation strategies. Furthermore, within surveillance systems, real-time vehicle detection equips security personnel with the means to swiftly identify potential security threats, enhancing overall situational awareness and safety. This paper contributes to this dynamic landscape by introducing an innovative architecture expressly crafted for vehicle detection. Leveraging state-of-the-art computer vision techniques and deep learning methodologies, this architecture addresses the evolving challenges posed by contemporary traffic scenarios. Through the enhancement of real-time vehicle detection capabilities, our aim is to fortify the foundations of safety, efficiency, and security within vehicular contexts. The paper includes a fusion of AlexNet and Darknet accompanied by a comparative analysis. It is expected to deliver unmatched dependability and accuracy in real-time traffic monitoring, setting a new standard for performance in this application. [ABSTRACT FROM AUTHOR]

In this paper, the optimal formation control issue with collision avoidance and connectivity assurance is investigated for a class of second-order uncertain nonlinear multi-agent systems. First, the neural networks are employed in order to deal with the unknown nonlinear dynamics of the system. Then, an optimal formation control scheme is developed in the framework of the identifier–actor–critic. By constructing a new performance metric function containing collision avoidance and connectivity constraints, it is demonstrated that asymptotic convergence of the tracking error can be achieved under the proposed control scheme. Finally, the effectiveness of the proposed control method is validated by the numerical simulation example. [ABSTRACT FROM AUTHOR]

Optimal portfolio selection—composing a set of stocks/assets that provide high yields/returns with a reasonable risk—has attracted investors and researchers for a long time. As a consequence, a variety of methods and techniques have been developed, spanning from purely mathematics ones to computational intelligence ones. In this paper, we introduce a method for optimal portfolio selection based on multi-objective evolutionary algorithms, specifically Nondominated Sorting Genetic Algorithm-II (NSGA-II), which tries to maximize the yield and minimize the risk, simultaneously. The system, named EvoFolio, has been experimented on stock datasets in a three-years time-frame and varying the configurations/specifics of NSGA-II operators. EvoFolio is an interactive genetic algorithm, i.e., users can provide their own insights and suggestions to the algorithm such that it takes into account users' preferences for some stocks. We have performed tests with optimizations occurring quarterly and monthly. The results show how EvoFolio can significantly reduce the risk of portfolios consisting only of stocks and obtain very high performance (in terms of return). Furthermore, considering the investor's preferences has proved to be very effective in the portfolio's composition and made it more attractive for end-users. We argue that EvoFolio can be effectively used by investors as a support tool for portfolio formation. [ABSTRACT FROM AUTHOR]

Assessing the strength of complex networks is crucial for evaluating their functionality and monitoring system security. Two primary perspectives for measuring network robustness are connectivity and controllability, which quantify how well a network maintains connectivity and controllability after experiencing complete attacks. However, conventional robustness calculation methods that use simulation attacks can be impractical for large-scale networks. Recent research has explored the use of deep learning methods for network robustness prediction, including PCR (short for predictor of controllability robustness), which uses a VGG (short for visual geometry group)-based convolutional neural network (CNN) and a linear filter. However, PCR has limitations, including inaccurate edge importance differentiation and fluctuating output curves. This paper presents ATTRP, an attention-based robustness predictor, which addresses these limitations. ATTRP uses a novel convolutional neural network with channel and spatial attention mechanisms to improve prediction accuracy and a robustness filter based on Savitzky–Golay smoothing to rectify the robustness curve. Experimental studies demonstrate that ATTRP outperforms other predictors, with much lower errors, while its computational speed is far superior to that of traditional attacking simulation. Overall, our research provides a more effective and efficient means of predicting network robustness, with potential applications in monitoring system security. [ABSTRACT FROM AUTHOR]

As a key component of rotating machinery, it is of great significance for the timely diagnosis of bearing weak faults. Stochastic resonance is widely used for its special signal enhancement pattern, and the combination of system parameters determines its actual output effect. Because the social mimic optimization algorithm has the advantages of few parameters, fast convergence speed and strong exploitation capability, it is used to optimize the parameters of stochastic resonance system in this paper. Aiming at the problem that it is easy to fall into local optimum during optimization, inspired by the learning habits of elite, an elite social mimic optimization (ESMO) algorithm is proposed. Its improvement mainly includes five parts: integration of learning efficiency, exchange learning, looking for successors, innovation and breakthrough, elimination mechanism. And its superiority is verified by the comparison of 29 standard benchmark functions and 10 other well-known optimization algorithms. Aiming at the discontinuity of optimization space, the concept of survival rate (surr) is proposed, and the effectiveness of the ESMO algorithm in optimizing discontinuous variables is analyzed and verified. Aiming at the disadvantage that stochastic resonance can only process small frequency signals, a bearing weak fault diagnosis method based on frequency exchange and parameter compensation stochastic resonance is proposed. To verify its application ability in other fault diagnosis methods, a classification and recognition method based on BP neural network is proposed. Finally, the effectiveness and superiority of the ESMO algorithm are verified by simulation signals and bearing experimental data. The above analysis results prove that the ESMO algorithm has certain scientific and engineering application value in the optimization field and engineering application. [ABSTRACT FROM AUTHOR]

In this paper, we primarily concentrate on the quasi-projective synchronization of master–slave discrete-time BAM neural networks. Without using the LMI and matrix measure methods, by applying discrete inequalities and the Lyapunov sequences, constructing a discrete inequality group, then solving the inequality group, two novel criteria to guarantee the quasi-projective synchronization are obtained for the considered networks. The controllers designed, the approach, and the results in the paper are fully novel, which advance the development of the study of synchronization of discrete neural networks. [ABSTRACT FROM AUTHOR]

In this paper, we present a novel approach to achieve finite-time synchronization (FTS) in a certain class of fractional-order complex-valued neural networks (CVNNs) containing reaction-diffusion terms. The proposed method uses intermittent control and provides a theoretical analysis to establish criteria for achieving FTS. This is achieved through new Lyapunov functions based on the proposed system, deriving inequalities in the complex domain. To realize FTS, the study designs complex-valued intermittent controllers for the targeted CVNNs relying solely on the information obtained from the controlled nodes. Moreover, an adaptive controller is introduced to effectively regulate the control gain, and the FTS of CVNNs is analyzed. The effectiveness of the proposed control strategies and derived results is demonstrated by numerical examples. [ABSTRACT FROM AUTHOR]

This paper introduces a cascade one proportional derivative incorporating filter (1PDf)-proportional integral (PI) controller abbreviated as c-1PDf-PI to deal effectively with the speed control issue of brushless DC (BLDC) motors. Two problems exist with implementing this controller such as iterated integral overflow and derivation-based chattering owing to the noise. The former is resolved by using an equivalent expression for the integral operation, while the latter is addressed by putting a first-order filter on the derivative term. To achieve the best performance from the controller, snake optimizer (SO) is fruitfully employed for optimizing the controller parameters without need for expert knowledge/interpretation. Here, a more reasonable cost function to assess the candidate solutions is also described. Simulations and laboratory experiments using DSP of TI TMS320F28335 are performed and the results are presented which show that the reference tracking performance, torque disturbance capability and robustness of the c-1PDf-PI controller have potential. These results are also contrasted by those offered by PI and 1PDf speed control schemes individually, affirming the superior performance of our proposal. As per the results, discussion and observation of this research, we stress that good performance and simplicity are salient advantages of the c-1PDf-PI controller, rendering it a good alternative over the complicated controller designs. [ABSTRACT FROM AUTHOR]

Few-shot semantic segmentation (FSS) is a challenging task that aims to segment new classes in query images with a few annotated support samples. One inherent challenge in FSS is the intra-class variation resulting from the limited availability of support samples and the diversity of query data. Current methods frequently employ prototype-support techniques to tackle this issue. However, it is important to note that a single support prototype shares limited commonalities with query features, and increases the difficulty of accurate segmentation. In this paper, we propose a lightweight and effective framework named Siamese Transformer (SiaT) with a mere 0.68M learnable parameters to enhance commonalities between prototypes and query features. The SiaT framework consists of two key modules: the Siamese Transformer Module (STM) and the Query Activation Module (QAM). The STM integrates two shared Transformer decoders for information propagation to generate two enhanced prototypes. One Transformer decoder facilitates the propagation of target-related information from the query to the original support prototype, while the other propagates foreground information from the support to the initial query prototype. The QAM utilizes the two enhanced prototypes from the STM as support information, engaging with query features through channel-wise allocation and concatenation, termed query activation. Moreover, SiaT showcases competitive performance on the widely used benchmarks PASCAL- 5 i and COCO- 20 i , which demonstrates its effectiveness in addressing the intra-class variation challenge within FSS tasks. [ABSTRACT FROM AUTHOR]

Feature selection (FS) is a crucial preprocessing step that aims to eliminate irrelevant and redundant features, reduce the dimensionality of the feature space, and enhance clustering efficiency and effectiveness. FS is categorized as NP-Hard due to the high number of existing solutions. Various metaheuristic methods have been developed to address the FS problem, yielding promising results. Particularly, particle swarm optimization (PSO), an evolutionary computing (EC) approach guided by swarm intelligence, has gained widespread adoption owing to its implementation simplicity and potential for global search. This paper analyzes several variants of PSO algorithms and introduces a new FS method called HPSO. The proposed approach utilizes an asynchronously adaptive inertia weight and an improved constriction factor. Additionally, it incorporates a chaotic map and a MAD fitness function with a feature count penalty to tackle the clustering FS problem. The efficiency of the developed method is evaluated against the genetic algorithm (GA) and well-known variants of PSO algorithms, including PSOs with fixed inertia weights, PSOs with improved inertia weights, PSOs with fixed constriction factors, PSOs with improved constriction factors, PSOs with adaptive inertia weights, and PSO's includes advanced learning exemplars and sophisticated structure topologies. This paper assesses two different reference text data sets, Reuters-21578 and Webkb. In comparison with competitive methods, the proposed HPSO method achieves higher clustering precision and selects a more informative feature set. [ABSTRACT FROM AUTHOR]

Gliomas are the most common and aggressive form of all brain tumors, with a median survival rate of fewer than two years, especially for the highest-grade glioma patient. Accurate and reproducible brain tumor segmentation is essential for an effective treatment plan and diagnosis to reduce the risk of further spread. Automated brain tumor segmentation is challenging because it can appear in the brain with variations in shape, size, and position from one patient to another. Several deep learning architectures have been created to handle automatic segmentation with good performance results on 3D MRI images. However, these architectures are generally large and require high hardware specifications and a large amount of memory and storage. This paper proposes a lightweight modified 3D UNet architecture with an outstanding performance level called Yaru3DFPN. The architecture is built based on the UNet. The block used is ResNet and is modified to use pre-activation strategies and GroupNormalization for batch normalization. In the expanding section, features are arranged into pyramid features. The final output is thresholded using the combining thresholding method. This architecture is light and fast. This proposal was tested using BraTS datasets with the highest dice performance of 80.90%, 86.27%, and 92.02% for ET, TC, and WT areas, respectively. This result outperformed all other comparative architectures and promised to be developed for clinical application. [ABSTRACT FROM AUTHOR]

Over the preceding decade, there has been a discernible surge in the prominence of artificial intelligence, marked by the development of various methodologies, among which deep learning emerges as a particularly auspicious technique. The captivating attribute of deep learning, characterised by its capacity to glean intricate feature representations from data, has served as a catalyst for pioneering approaches and methodologies spanning a multitude of domains. In the face of the burgeoning exponential growth in digital medical image data, the exigency for adept image compression methodologies has become increasingly pronounced. These methodologies are designed to preserve bandwidth and storage resources, thereby ensuring the seamless and efficient transmission of data within medical applications. The critical nature of medical image compression accentuates the imperative to confront the challenges precipitated by the escalating deluge of medical image data. This review paper undertakes a comprehensive examination of medical image compression, with a predominant focus on sophisticated, research-driven deep learning techniques. It delves into a spectrum of approaches, encompassing the amalgamation of deep learning with conventional compression algorithms and the application of deep learning to enhance compression quality. Additionally, the review endeavours to explicate these fundamental concepts, elucidating their inherent characteristics, merits, and limitations. [ABSTRACT FROM AUTHOR]

The adversary makes malicious samples capable of triggering erroneous judgments in deep learning models by introducing imperceptible perturbations to the original benign texts. These malicious samples are referred to as adversarial texts. The exploration of adversarial text generation methods not only facilitates our understanding of the robustness of mainstream deep neural networks against such adversarial attacks but also aids in developing appropriate defensive strategies. Nevertheless, the mainstream research on textual adversarial attacks has mainly focused on attack effectiveness, overlooking the associated attack cost. For real-world attacks, considerations such as the time cost, material cost, manpower cost, and various constraints are also crucial. In this paper, we propose a low-cost adversarial text generation method based on the universal strategy in the black-box attack scenario, Universal Chinese Text Tricker (UCTT), for tendency classification on Chinese texts. UCTT is both text-independent and model-independent, which markedly reduces its attack cost. Instead of crafting adversarial texts for a specific text, UCTT generates universal perturbations based on a universal word substitution list, which is applicable to any data in tendency classification datasets. Given a perturbation rate, we can use the word list to craft adversarial texts by simple substitutions without accessing the target model. In the framework of adversarial text generation based on word importance, UCTT utilizes count, arithmetic progression, linear normalization, and nonlinear normalization to calculate the scores of the important words in the dataset and then computes the candidate word frequencies, which in turn constructs the universal word substitution list. Compared with other black-box methods, the experimental results on real-world tendency classification datasets show that UCTT exhibits an effective attack capability while significantly reducing the attack cost. Compared to the powerful baseline we designed that exceeds the SOTA, UCTT improves the efficiency of adversarial text generation by up to a factor of 7 without accessing the target model. In addition to demonstrating excellent attack performance on mainstream models, UCTT is also capable of attacking the powerful ChatGPT in the physical world, which cannot be directly attacked by traditional adversarial text generation methods due to the hard labels produced by the target model. [ABSTRACT FROM AUTHOR]

Multi-Object Tracking, also known as Multi-Target Tracking, is an important area of computer vision with various applications in different domains. The advent of deep learning has had a profound impact on this field, forcing researchers to explore innovative avenues. Deep learning methods have become the cornerstone of today's state-of-the-art solutions, consistently delivering exceptional tracking results. However, the significant computational demands of deep learning models require powerful hardware resources that do not always match real-time tracking requirements, limiting their practical applicability in real-world scenarios. Thus, there is an imperative to strike a balance by merging robust deep learning strategies with conventional approaches to enable more accessible, cost-effective solutions that meet real-time requirements. This paper embarks on this endeavor by presenting a hybrid strategy for real-time multi-target tracking. It effectively combines a classical optical flow algorithm with a deep learning architecture tailored for human crowd tracking systems. This hybrid approach achieves a commendable balance between tracking accuracy and computational efficiency. The proposed architecture, subjected to extensive experimentation in various settings, demonstrated notable results, achieving a Mean Object Tracking Accuracy (MOTA) of 0.608. This level of performance placed it as the highest ranking solution on the MOT15 benchmark, surpassing the state-of-the-art benchmark of 0.549, and consistently ranked among the superior models on the MOT17 and MOT20 benchmarks. Additionally, the incorporation of the optical flow phase resulted in a substantial reduction in processing time, nearly halving the duration, while simultaneously maintaining accuracy levels comparable to established techniques. [ABSTRACT FROM AUTHOR]

Source code is an intermediary through which humans communicate with computer systems. It contains a large amount of domain knowledge which can be learned by statistical models. Furthermore, this knowledge can be used to build software engineering tools. We find that the functionality of the source code depends on the programming language-specific token which build the base structure, while identifiers provide natural language information. On this basis, we found that the knowledge in the source code can be sufficiently learned more when modeling the source code in bimodal. This paper presents the bimodal composition language model (BCLM) for source code modeling and representation. We analyze the effectiveness of bimodal modeling, and the results show that the bimodal approach has great potential for source code modeling and program comprehension. [ABSTRACT FROM AUTHOR]

Recent advancements in the industrial revolution and artificial intelligence have aided in the development of novel approaches that have considerable potential for increasing the efficiency of traffic networks. Emerging concepts of autonomous driving and machine learning can be incorporated intelligently for improving traffic operations and control. In the coming years, traffic composition can vary in terms of autonomous vehicle (AV) penetration rate, which can lead to a heterogeneous traffic environment. Traffic control methods for such complex networks need to be designed effectively for accommodating the positive effects of AV implementation without compromising the safety and level of service in the presence of regular vehicles (RVs). An intelligent-based optimization framework for traffic signal control under heterogeneous AV-based traffic is proposed in this paper. This framework utilizes state-of-the-art machine-learning approaches to represent and design different components of the optimization process of cycle length. Further, SHapley Additive exPlanations (SHAP) is used to enhance model interpretability. The proposed optimization framework improves performance under congested traffic conditions compared with that of the conventional optimization methods. Compared to pure RV-based traffic, the penetration rates of 25%, 50%, and 100% can decrease optimized cycle lengths by 26%, 39%, and 53%, respectively, which can result in delay reductions of approximately 18%, 31%, and 56%, respectively. The proposed framework when applied in an adaptive-based manner can help in the generalization of controlling existing signalized intersections with the gradual penetration of AVs without the extra infrastructure or specific operational and connectivity capabilities of the involved vehicles. [ABSTRACT FROM AUTHOR]

Implicit discourse relation recognition (IDRR) is to detect and classify relation sense between two text segments without an explicit connective. Existing neural network models learn a semantic representation for each argument from its compositional words and classify discourse relation by the interactive semantic representation of an argument pair. As the basic English unit, a word only carries a simple and limited meaning that may be incorrectly interpreted in an argument without considering its syntactic context. This motivates us to explore whether we can learn an argument’s semantic representation from another language unit phrase which usually contains full contextual meaning. We also argue that some semantic connection in between phrase pairs can be further exploited to infer the discourse relation between arguments. In this paper, we propose an Attentive Phrase Interaction Learning (APIL) model to parse and encode the interactive phrase structure for the IDRR task. In APIL, we propose a minimum subtree algorithm to obtain the phrase sequence of input arguments from its constituency syntax tree. We also design an Attentive Matching Network to learn the representation for each phrase from both arguments’ semantic context and words’ linguistic evidence. Furthermore, we propose a Phrase Inference Network to encode the linguistic relations of phrase pairs as semantic connections for relation inference. Experiments on the PDTB corpus show that our APIL model outperforms the state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Age-invariant face recognition (AIFR) is a significant research task in general face recognition, and it aims at eliminating gradual discordance of individual’s facial appearance caused by aging process. Previous discriminative methods decompose facial components over one-dimensional feature vectors which overlooks critical facial information and hypothesize linear relationships over aging process which is inadequate to describe complex correlations. In this paper, we propose an enhanced AIFR model, namely age-invariant face network (AFN), to eliminate the discrepancy of aging process over facial semblance. Specifically, we propose attentive factorization module (AFM) leveraging attention mechanism to decompose facial features into identity-related features and age-related features in two-dimensional space on both local and contextual levels. We take both linear and nonlinear correlation analyses into account for a better reflection of aging/rejuvenation process and hence propose a hybrid correlation regularizer (HCR) to supervise the decorrelation between factorized features. Both identity features and age features are supervised simultaneously in a multi-task learning framework where only identity features are used in test phase for evaluation of AIFR performance. Experiments across common cross-age datasets (e.g., FG-Net, CACD-VS, CALFW, AgeDB-30) show the effectiveness of proposed AFN. Further, our proposed AFN is validated over LFW dataset to demonstrate its effectiveness on general face recognition task. [ABSTRACT FROM AUTHOR]

In 2022, the World Health Organization declared an outbreak of monkeypox, a viral zoonotic disease. With time, the number of infections with this disease began to increase in most countries. A human can contract monkeypox by direct contact with an infected human, or even by contact with animals. In this paper, a diagnostic model for early detection of monkeypox infection based on artificial intelligence methods is proposed. The proposed method is based on training the artificial neural network (ANN) with the adaptive artificial bee colony algorithm for the classification problem. In the study, the ABC algorithm was preferred instead of classical training algorithms for ANN because of its effectiveness in numerical optimization problem solutions. The ABC algorithm consists of food and limit parameters and three procedures: employed, onlooker and scout bee. In the algorithm standard, artificial onlooker bees are produced as much as the number of artificially employed bees and an equal number of limit values are assigned for all food sources. In the advanced adaptive design, different numbers of artificial onlooker bees are used in each cycle, and the limit numbers are updated. For effective exploitation, onlooker bees tend toward more successful solutions than the average fitness value of the solutions, and limit numbers are updated according to the fitness values of the solutions for efficient exploration. The performance of the proposed method was investigated on CEC 2019 test suites as examples of numerical optimization problems. Then, the system was trained and tested on a dataset representing the clinical symptoms of monkeypox infection. The dataset consists of 240 suspected cases, 120 of which are infected and 120 typical cases. The proposed model's results were compared with those of ten other machine learning models trained on the same dataset. The deep learning model achieved the best result with an accuracy of 75%. It was followed by the random forest model with an accuracy of 71.1%, while the proposed model came third with an accuracy of 71%. [ABSTRACT FROM AUTHOR]

Scene recognition has been the foundation of research in computer vision fields. Because scene images typically are composed of specific regions distributed in some layout, so modeling layouts of various scenes is a key clue for scene recognition. Existing methods usually require an additional stream to detect regions for subsequent modeling, which accumulate errors and may miss important information. Meanwhile, they use manual features to model relations between regions, which weakens the representation ability of layouts. In this paper, we propose a single-stream adaptive scene layout modeling approach based on a layout modeling module (LMM), which constructs layouts without additional detection streams and adaptively captures the relations to take advantage of graph attention network. LMM is directly concatenated to a convolutional neural network, where each pixel of the activation maps of the last convolutional layer is defined as a region that is the initial input node of the LMM. LMM first models the layout of each region, and then uses all regions with layout information to model the entire scene. Layout relations are encoded as edges, which are automatically analyzed according to region co-occurrence and relative position. Our work can be understood as optimizing features of the activation maps from a scene layout modeling perspective for scene recognition. Experimental results on MIT67, SUN397, and Places365 show that our single-stream model achieves competitive performance. [ABSTRACT FROM AUTHOR]

To enhance the safety of direct physical interaction between humans and lower-limb exoskeletons, the actuator of the lower-limb exoskeleton should possess compliance, force controllability, and back-drivability. In this paper, a novel compact compliant force control actuator with worm gear and gear rack as transmission modes is designed to provide flexible power to the lower-limb exoskeleton. The worm gear is utilized to amplify the torque generated by the motor in the compact compliant actuator, while the gear rack is established to improve the bearing capacity. Yet, the friction is occurred between the worm gear, rack and other mechanical configurations which may lead to non-negligible force control error of the compliant actuator. A noise-tolerant zeroing neural network controller is proposed to suppress noises. Additionally, theoretical proofs are provided for the convergence performance of the noise-tolerant zeroing neural network controller, as well as the suppression of constant noise, linear disturbances, and bounded random noise. The proposed noise-tolerant zeroing neural network controller is validated on the compliant actuator through numerical simulations, experimental results, and comparison experiments, demonstrating its effectiveness in suppressing various noise and improving the convergence, stability, and robustness of the compliant actuator system. Additionally, the controller is further verified through walking experiments conducted on a knee robot. [ABSTRACT FROM AUTHOR]

Accurate forecasting of electricity generation from renewable energy sources is crucial for the operation, planning and management of smart grids. For reliable planning and operation of photovoltaic (PV) systems in grid-connected or islanded utilities, an hourly day-ahead forecast of PV output is critical. The forecast of PV power can be done indirectly by estimating solar irradiance. For forecasting day-ahead hourly global horizontal irradiance (GHI), two forecasting models with different multivariate inputs are proposed in this paper, and the results are compared. These models use a hybrid algorithm of discrete wavelet decomposition and bidirectional long short-term memory (BILSTM). The inputs of the first model contain GHI and weather type data. The other model allows for observation of the effect of meteorological values including GHI, temperature, humidity, wind speed, and weather type data. The forecasting performance of deep learning algorithms which contain recurrent neural network (RNN), long short-term memory (LSTM), and BILSTM algorithms for day ahead hourly solar irradiance forecasting problems are also compared. To evaluate the performance of proposed models, two datasets are used for Model 1 and one dataset is used for Model 2. An experiment is also done to demonstrate that the proposed Model 1 is applicable in datasets collected in the vicinity of the city of Trabzon. On the other hand, BILSTM algorithm outperforms RNN and LSTM algorithms. It is seen that the test successes of both proposed models are better than the results given in the literature. [ABSTRACT FROM AUTHOR]

Images captured under low-light conditions often suffer from low contrast, high noise, and uneven brightness due to nightlight, backlight, and shadow. These challenges make it difficult to use them as high-quality inputs for visual tasks. Existing low-light enhancement methods tend to increase overall image brightness, which can cause overexposure of normal-light areas after enhancement. To solve this problem, this paper proposes an Uneven Dark Vision Network (UDVN) that consists of two sub-networks. The Luminance Domain Network (LDN) uses Direction-aware Spatial Context (DSC) and Feature Enhancement Module (FEM) to segment different light regions in the image and output the luminance domain mask. Guided by this mask, the Light Enhancement Network (LEN) uses the Cross-Domain Transformation Residual block (CDTR) to adaptively illuminate different regions with various lights. We also introduce a new region loss function to constrain the LEN to better enhance the quality of different light regions. In addition, we have constructed a new low-light synthesis dataset (UDL) that is larger, more diverse, and includes uneven lighting states in the real world. Extensive experiments on several benchmark datasets demonstrate that our proposed method is highly competitive with state-of-the-art (SOTA) methods. Specifically, it outperforms other methods in light recovery and detail preservation when processing uneven low-light images. The UDL dataset is publicly available at: https://github.com/YuhangLi-li/UDVN. [ABSTRACT FROM AUTHOR]

This paper focuses on the prescribed-time cluster synchronization of coupled inertial neural networks. By variable transformation, the coupled inertial neural networks are converted into high-order systems. An effective prescribed-time control applicable to high-order systems is introduced by virtue of a time-varying scaling function. Moreover, a dimension-lifting approach is utilized to derive sufficient criteria for achieving prescribed-time cluster synchronization under the proposed control. Compared with existing Lyapunov–Krasovskii functional methods, the criteria obtained in this paper are in form of low-dimensional linear matrix inequalities and therefore can be easily verified. A numerical example is provided to demonstrate the effectiveness of proposed results. [ABSTRACT FROM AUTHOR]

The water quality index (WQI) serves as a global representation of river water quality (WQ). Existing studies related to the WQI have mainly focused on two aspects: (i) a WQI point estimation using multiple WQ inputs; and (ii) a one-step-ahead WQI forecasting with datasets of lower temporal resolution. These approaches, however, are limited in their ability to forecast future trends of the WQI for timely and prompt responses to pollution events. In this study, the deep learning algorithms, namely the long short-term memory (LSTM) and the gated recurrent unit (GRU), were selected for direct multi-step-ahead WQI forecasting. To enhance the capability of the models in capturing their temporal patterns, the input signal was pre-decomposed using the empirical mode decomposition (EMD) and variational mode decomposition (VMD) into several intrinsic mode functions (IMFs). The characteristics of these IMFs were then analyzed and used to ease model learning on capturing their temporal patterns. Our study shows that the selection of signal decomposition strategies significantly impact the model performance. Both deep learning algorithms offered comparable performances, with the VMD-LSTM exhibiting the lowest prediction errors (MAPE = 1.9237%) and the highest Kling–Gupta efficiency (KGE = 0.6761) over a two-month test period. To the best of our knowledge, this is the first paper that applies a direct forecast approach for the multi-step-ahead WQI forecasting, using the IMFs obtained from the EMD and VMD decompositions. The performance of the model was evaluated through a rolling forward setup to ensure its consistency across different test periods. The proposed modeling framework holds the potential to assist policymakers and stakeholders in decision-making, particularly in planning remedies and efficient water resource management. [ABSTRACT FROM AUTHOR]

Nowadays, landing gears used in aircraft are the basic systems that carry the weight of aircraft systems and provide them to move on the ground. During the landing of aircraft, big and sudden vibrations occur as soon as the wheels come into contact with the runway. This situation negatively affects human safety, comfort and also the life of the landing gear system. Therefore, it is important to keep this vibration at permissible levels. This paper presents an improved vibration control method that controls the amplitudes of vibrations of landing systems. This paper presents proposed vibration control method that controls the angular position variables of the landing gear mechanism to control the vibrations occurring during landing. For different runway conditions and 200 km/h landing speed, the angular joint variables of the landing gear mechanism links are controlled using standard controller and a proposed neural controller structure. The simulation results are shown that the proposed neural controllers have better performance at vibration control of landing systems during landing. [ABSTRACT FROM AUTHOR]

Question generation (QG) aims to generate natural questions from relevant input. Existing state-of-the-art QG approaches primarily leverage pre-trained language models (PLMs) to encode the deep semantics within the input. Meanwhile, studies show that the input's dependency parse tree (referred to as syntactic information) is promising in improving NLP-oriented tasks. However, how to incorporate syntactic information in PLMs to guide a QG process effectively still needs to be settled. This paper introduces a syntax-guided sentence-level QG model based on prompt learning. Specifically, we model the syntactic information by utilizing soft prompt learning, jointly considering the syntactic information from a constructed dependency parse graph and PLM to guide question generation. We conduct experiments on two benchmark datasets, SQuAD1.1 and MS MARCO. Experiment results show that our model exceeded both automatic and human evaluation metrics compared with mainstream approaches. Moreover, our case study shows that the model can generate more fluent questions with richer information. [ABSTRACT FROM AUTHOR]

Small object detection is widely used in a variety of fields such as automatic driving, UAV-based object detection, and aerial image detection. However, small objects carry limited information, making it difficult for detectors to detect small objects. In recent years, the development of deep learning has significantly improved the performance of small object detection. This paper provides a comprehensive review to help further the development of small target detection. We summarize the challenges related to small object detection and analyze solutions to these challenges in existing works, including integrating the feature at different layers, enriching available information, balancing the number of positive and negative samples for small objects, and increasing sufficient small object instances. We discuss related methods developed in three application areas, including automatic driving, UAV search and rescue, and aerial image detection. In addition, we thoroughly analyze the performance of typical small object detection methods on popular datasets. Finally, based on the comprehensive review of small object detection methods, we point out possible research directions for future studies. [ABSTRACT FROM AUTHOR]

Representation learning serves as a crucial link between knowledge graphs and neural models. Knowledge graphs are typical symbolic models and require representation learning to obtain vector representations of entities and relationships. Existing efforts focus more on the structural characteristics of knowledge itself, such as the connections between entities and relationships under the same type of knowledge, neglecting the potential value of commonsense knowledge in guiding and constraining. In fact, commonsense knowledge implies the belonging tendency of entities in triplets. In this paper, we propose a novel knowledge representation learning model, CoolGust, which is guided and constrained by commonsense knowledge and can effectively utilize commonsense knowledge to seamlessly guide the belonging relationships of entities and enhance the model performance. Commonsense knowledge is essential for guiding humans to make wise decisions in unknown scenarios. We find that by utilizing commonsense knowledge as guides and constraints for entities, hidden knowledge entanglement structures can be formed in complex network applications, thereby constructing decisions consistent with situational logic. To verify the effectiveness of our model, we conducted experimental verification of two tasks, link prediction, and triple classification, on three public datasets. The experimental results demonstrate the effectiveness and advancedness of our proposed method. [ABSTRACT FROM AUTHOR]

Deep learning integration in cancer diagnosis enhances accuracy and diagnosis speed which helps clinical decision-making and improves health outcomes. Despite all these benefits in cancer diagnosis, the present AI models in urology cancer diagnosis have not been sufficiently reviewed systematically. This paper reviews the artificial intelligence approaches used in cancer diagnosis, prediction, and treatment of urology cancer. AI models and their applications in urology subspecialties are evaluated and discussed. The Scopus, Microsoft Academic and PubMed/MEDLINE databases were searched in November 2022 using the terms "artificial intelligence", "neural network", "machine learning," or "deep learning" combined with the phrase "urology cancers". The search was limited to publications published within the previous 20 years to identify cutting-edge deep-learning applications published in English. Irrelevant review articles and publications were eliminated. The included research involves two kinds of research analysis: quantitative and qualitative. 48 articles were included in this survey. 25 studies proposed several approaches for prostate cancers, while 15 were for bladder cancers. 8 studies discussed renal cell carcinoma and kidney cancer. The models presented to detect urology cancers have achieved high detection accuracy (77–95%). Deep learning approaches that use convolutional neural networks have achieved the highest accuracy among other techniques. Although it is still progressing, the development of AI models for urology cancer detection, prediction, and therapy has shown significant promise. Additional research is required to employ more extensive, higher-quality, and more recent datasets to the clinical performance of the proposed AI models in urology cancer applications. [ABSTRACT FROM AUTHOR]

With the introduction of ChatGPT, the public's perception of AI-generated content has begun to reshape. Artificial intelligence has significantly reduced the barrier to entry for non-professionals in creative endeavors, enhancing the efficiency of content creation. Recent advancements have seen significant improvements in the quality of symbolic music generation, which is enabled by the use of modern generative algorithms to extract patterns implicit in a piece of music based on rule constraints or a musical corpus. Nevertheless, existing literature reviews tend to present a conventional and conservative perspective on future development trajectories, with a notable absence of thorough benchmarking of generative models. This paper provides a survey and analysis of recent intelligent music generation techniques, outlining their respective characteristics and discussing existing methods for evaluation. Additionally, the paper compares the different characteristics of music generation techniques in the East and West as well as analysing the field's development prospects. [ABSTRACT FROM AUTHOR]

Brain tumor detection and classification are crucial steps in evaluating life-threatening abnormal tissues to provide appropriate treatment plans. For clinical assessment, Magnetic resonance imaging (MRI) is normally used because of its excellent quality and lack of ionizing radiation. However, as the volume of the data grows, manual processing of MRI images becomes expensive, time-taking, and error prone. Also, traditional automated detection systems struggle to handle complex image patterns, leading to reduced classification accuracy. So, this paper designs a reliable and effective brain tumor detection mechanism as a solution to these problems. The proposed "Vision Transformer with Attention and Linear Transformation module (VITALT)" system is a combination of modules such as Vision Transformer (ViT), Split bidirectional feature pyramid network (S-BiFPN), linear transformation module (LTM) and soft-quantization that effectively extracts features from complex brain structures. At first, to mitigate the training inaccuracies developed by dimension and quality constraints, the preprocessing steps such as resizing and normalization are executed. The preprocessed images are divided into number of patches and embedded into high-dimensional vector to provide more compact image representation. Subsequently, the global and local features in the image are captured through ViT module by learning the relationship between image patches. The multi-scale spatial features formed are then fused using S-BiFPN to increase the accuracy of prediction. By using LTM to improve the linear expression capability of the design, the characteristics that are most important for the classification of brain tumors are discovered. Also, soft quantization is used to minimize memory footprint and minimize quantization errors in detection. Finally, the head module with set of fully connected layers accurately classifies different classes of brain tumors. The experimental analysis conducted using four different benchmark brain tumor datasets shows the viability and reliability of the suggested VITALT system in predicting brain tumors, as measured by multiple evaluation metrics. The proposed system achieves classification accuracy of 99.08% for Dataset A, 98.97% for Dataset B, 98.82% for Dataset C and 99.15% for Dataset D. A high level of classification accuracy attained by the suggested system highlights its potential in medical imaging applications and its ability to contribute to improved surgical treatments. [ABSTRACT FROM AUTHOR]