Search

In the field of paper conservation and archival research, identifying the raw materials of paper is important to elucidate its history and culture. As the most basic element of the raw materials for paper, fibers have not been sufficiently investigated. In this study, we propose a nondestructive method for estimating paper fibers from macro photographs ( $4000\times3000$ pixels) captured using a digital camera. The proposed method consists of background patch (500 pixels per side) detection (BPD), wherein background regions with no text are identified; patch fiber classification (PFC), wherein background patches obtained after BPD are analyzed for fiber classification; and paper fiber estimation (PFE), wherein macro images obtained after PFC are analyzed for fiber estimation. BPD and PFC are employed to perform patch-based classification on segmented macro images, which are reconstructed during PFE to obtain the final fiber estimation results. We performed experiments using 1337 macro images (64176 patches) to evaluate the fiber estimation accuracy for kozo, mitsumata, and gampi via three-fold cross-validation. The average fiber classification accuracy for patch images was observed to be 79.1%; accordingly, the average fiber estimation accuracy for macro images was 85.8%. Experimental results indicated that PFE can be realized in a nondestructive manner on macro images of paper captured using a digital camera.

In the era of Industry 4.0, digital transformation has spurred the swift advancement of technologies, including intelligent predictive maintenance scheduling, prognostics and health management. The accurate prediction of remaining useful life plays a crucial role in these technologies as it extends power equipment’s safe operational duration and decreases the maintenance costs associated with unforeseen shutdowns. Also, the increased accessibility of data for monitoring system conditions has paved the way for the more immense adoption of machine learning models in prognostics and health management for power transformers. At the moment, with the ever-increasing demand for electricity, there is a corresponding increase in the degradation processes of power transformers. Transformers insulation system and more importantly, the paper insulation happens to be the principal part where the degradation is prominent. Therefore, an accurate prediction of the insulating paper condition through its degree of polymerization is required to guarantee the reliability of power transformers. In this regard, the predictions, reliability, and health monitoring of this power equipment can be actualized by modeling the degradation of transformer insulation paper through several machine learning frameworks. In this view, this review paper has been drafted not only to serve as a guide for researchers interested in the fields of transformer insulation system fault prognosis but also to offer insights into potential research directions as existing literature in modeling and evaluating transformer paper insulation is presented.

Evaluating the long-term influence of academic papers in the early stage after publication is crucial for research management and decision-making. Although the total number of citations has been considered as a valid indicator to measure the academic influence of papers, there are still some bottleneck problems in predicting future citation counts. Firstly, it remains a challenging task to select the reasonable features due to the diversity of model features. Secondly, some prediction models are overly complex and their defects may cause prediction bias or even errors. Finally, the most important thing is that very few models possess long-term predictive ability. This paper proposes a prediction model that can provide an early evaluation of the long-term citations, or the long-term influence, of academic papers. The model uses features available immediately after publication or after a certain period of time. Taking academic papers in Information Science and Library Science as our experimental subject, we train the model using three machine learning algorithms and find that the artificial neural network performs satisfactorily. Our findings may offer guidance for research management and decision-making.

This special section of the IEEE Open Journal of Circuits and Systems (OJCAS) aims to highlight a selection of papers from 2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS). Due to COVID-19 and travel restrictions, APCCAS 2022 was organised as a hybrid conference during 11 - 13 November 2022 in Shenzhen China. As the regional flagship conference of IEEE Circuits and Systems Society, APCCAS 2022 was driven by the theme “Building a Fully-connected AIoT World” to emphasize the CAS Society’s potential for finding multidisciplinary solutions to societal and industrial challenges. The papers in this special issue were selected from a comprehensive list of papers presented in the sessions of APCCAS 2022.

During the past two decades, a number of two-terminal switching devices have been demonstrated in the literature. They typically exhibit hysteric behavior in the current-to-voltage characteristics. These devices have often been also referred to as memristive devices. Their capacity to switch and exhibit electrical hysteresis has made them well-suited for applications such as data storage, in-memory computing, and in-sensor computing or in-memory sensing. The aim of this perspective paper is to is twofold. Firstly, it seeks to provide a comprehensive examination of the existing research findings in the field and engage in a critical discussion regarding the potential for the development of new non-Von-Neumann computing machines that can seamlessly integrate sensing and computing within memory units. Secondly, this paper aims to demonstrate the practical application of such an innovative approach in the realm of cancer medicine. Specifically, it explores the modern concept of employing multiple cancer markers simultaneously to enhance the efficiency of diagnostic processes in cancer medicine.

The Editorial Board of the IEEE Open Access Journal of Power and Energy (OAJPE) would like to recognize the following best papers selected from all papers published between October 1, 2020 and September 30, 2023 in OAJPE.

Welcome to Volume 4 of IEEE Journal of Microwaves! In this issue, we bring you 11 quality papers: two invited and nine regular submissions. We also celebrate six of our Outstanding Reviewers of 2023 and announce our IEEE Journal of Microwaves 2022 Best Paper prize. Our author and reader pools continue to expand and our usage counts per article published now exceed 1530. All of our 232 published articles from 2021 to 2023 now appear on Clarivate's Web of Science and we are looking forward to possibly receiving an impact factor this July. Remember to take a look at our solicitation for papers that can fit into our Microwaves in Climate Change special issue, which we intend to release before the end of 2024.

The article announces the IEEE International Symposium on Technology and Society, scheduled from September 18-20 in Puebla, Mexico, focusing on the Social Implications of Artificial Intelligence (AI), inviting submissions from academia, industry, and government to explore AI's impacts.

This research proposes an inkjet printed dual-band dual-sense circularly polarized antenna using CPW-feeding on PET substrate. The antenna is designed and optimized using ANSYS HFSS, which operates at 4.01 GHz - 5.05 GHz (22.96%) and 6.23 GHz - 7.58 GHz (19.55%) with a return loss of $0.44\lambda $ x $0.52\lambda $ x $0.002\lambda $ , where $\lambda $ is the free-space wavelength at the lowest circular polarization frequency. Measurement of the fabricated version shows good agreement with the simulated version. To the best of author’s knowledge, this proposed design is the first circularly polarized antenna integrating flexibility with dual-band and dual-sensing characteristics.

He is author of the books "Artificial Neural Networks for Modelling and Control of Non-linear Systems" (Springer) and "Least Squares Support Vector Machines" (World Scientific), co-author of the book "Cellular Neural Networks, Multi-Scroll Chaos and Synchronization" (World Scientific) and editor of the books "Nonlinear Modeling: Advanced Black-Box Techniques" (Springer), "Advances in Learning Theory: Methods, Models and Applications" (IOS Press) and "Regularization, Optimization, Kernels, and Support Vector Machines" (Chapman & Hall/CRC). [ABSTRACT FROM AUTHOR]

This issue of IEEE Vehicular Technology Magazine includes our first invited paper. In the next issues, we plan to publish a series of invited papers on selected topics of high interest that should drive and become a reference for the community. The invited papers will provide a vision on the evolution, open challenges, and trends for technologies of high-expected impact. [ABSTRACT FROM AUTHOR]

Recently, evolutionary computation (EC) has experienced significant advancements due to the integration of machine learning, distributed computing, and big data technologies. These developments have led to new research avenues in EC, such as distributed EC and surrogate-assisted EC. While these advancements have greatly enhanced the performance and applicability of EC, they have also raised concerns regarding privacy leakages, specifically the disclosure of optimal results and surrogate models. Consequently, the combination of evolutionary computation and privacy protection becomes an increasing necessity. However, a comprehensive exploration of privacy concerns in evolutionary computation is currently lacking, particularly in terms of identifying the object, motivation, position, and method of privacy protection. To address this gap, this paper aims to discuss three typical optimization paradigms, namely, centralized optimization, distributed optimization, and data-driven optimization, to characterize optimization modes of evolutionary computation and proposes BOOM (i.e., oBject, mOtivation, pOsition, and Method) to sort out privacy concerns related to evolutionary computation. In particular, the centralized optimization paradigm allows clients to outsource optimization problems to a centralized server and obtain optimization solutions from the server. The distributed optimization paradigm exploits the storage and computational power of distributed devices to solve optimization problems. On the other hand, the data-driven optimization paradigm utilizes historical data to address optimization problems without explicit objective functions. Within each of these paradigms, BOOM is used to characterize the object and motivation of privacy protection. Furthermore, this paper discuss the potential privacy-preserving technologies that strike a balance between optimization performance and privacy guarantees. Finally, this paper outlines several new research directions for privacy-preserving evolutionary computation. [ABSTRACT FROM AUTHOR]

This paper proposes an algorithm called Forward Composition Propagation (FCP) to explain the predictions of feed-forward neural networks operating on structured classification problems. In the proposed FCP algorithm, each neuron is described by a composition vector indicating the role of each problem feature in that neuron. Composition vectors are initialized using a given input instance and subsequently propagated through the whole network until reaching the output layer. The sign of each composition value indicates whether the corresponding feature excites or inhibits the neuron, while the absolute value quantifies its impact. The FCP algorithm is executed on a post-hoc basis, i.e., once the learning process is completed. Aiming to illustrate the FCP algorithm, this paper develops a case study concerning bias detection in a fairness problem in which the ground truth is known. The simulation results show that the composition values closely align with the expected behavior of protected features. The source code and supplementary material for this paper are available at https://github.com/igraugar/fcp. [ABSTRACT FROM AUTHOR]

This paper studies simultaneous scheduling of production and material transfer that arises in the semiconductor photolithography area. In particular, the right reticle and right job both need to be present to process the job. Jobs are transferred by a material handling system that employees a fleet of vehicles. Reticles serving as an auxiliary resource are also transferred from one place to another by a different set of vehicles. This intricate scheduling challenge, encompassing jobs, reticles, machines, and two distinct sets of vehicles, is explored here for the first time. The paper introduces a multi-stage methodology that involves relaxation, a constructive heuristic, constraint programming, and a warm-start approach to address this complex problem. [ABSTRACT FROM AUTHOR]

A scheduling problem for parallel batch processing machines (BPMs) with jobs having unequal ready times in semiconductor wafer fabrication facilities (wafer fabs) is studied in this paper. A blended objective function combining the total weighted tardiness (TWT) and the total electricity cost (TEC) under a time-of-use (TOU) tariff is considered. A genetic programming (GP) procedure is designed to automatically discover priority indices for a heuristic scheduling framework. Results of computational experiments are reported that demonstrate that the learned priority indices lead to high-quality schedules in a short amount of computing time. [ABSTRACT FROM AUTHOR]

Speaker recognition systems (SRSs) are commonly used for biometric identification. However, these systems are vulnerable to adversarial attacks. Several defenses have been proposed but they require high costs in terms of additional data and computational resources to ensure robustness. To address these issues, this paper proposes a low-cost input reconstruction defense method called adaptive F-ratio-based partial masking (AFPM), which utilizes a robust feature extraction process to guarantee high defensibility. The underlying distribution of non-robust features is explored and filtered out by partial masking (PM), which helps maintain a low defense construction cost. An F-ratio-based PM (FPM) defense strategy is proposed by integrating the F-ratio, which reflects the weight of each frequency band for distinguishing between speakers, to balance classification accuracy and defensiveness. AFPM, which introduces an adaptive threshold calculation algorithm to FPM, is proposed to achieve further improved defensiveness and flexibility. Comparative experimental results show that AFPM is low-cost, highly defensive and universal. The construction process of AFPM does not involve training and its implementation does not require the protected SRSs to be retrained, only fine-tuned. While maintaining the classification accuracy at 99.42%, the average defense capability of AFPM against five white-box adaptive attacks is 90.89%, which is 9.23% better than that of the low-cost input reconstruction defense method and 3.77% better than that of the high-cost Parallel WaveGAN (PWG) defense approach. Against grey- and black-box adaptive attacks, FAKEBOB and Kenansville, AFPM reaches maximum defense effects of 96.01% and 74.49%, respectively, surpassing PWG by 4.5% and 65.82%. Furthermore, AFPM is universal and capable of protecting various SRSs against different attack strengths. [ABSTRACT FROM AUTHOR]

This paper investigates the adaptive control for a class of uncertain nonlinear systems under denial-of-service (DoS) attacks. We analyze the closed-loop system stability under DoS attacks in terms of attack duration, attack frequency and resting time duration respectively. Three scenarios of DoS attacks against the system stability are considered. Firstly, it is shown that if the duration of each attack is less than a given constant, asymptotical convergence of system output is still preserved. Secondly, if the bounds on the frequency and duration of attacks with respect to overall intervals meet certain conditions, the proposed event-triggered control scheme guarantees that all the closed-loop signals are globally bounded and the stabilization error converges to a ball with a radius arbitrarily small. Thirdly, if resting time duration meets certain conditions after an arbitrarily long attack, closed-loop boundedness is still preserved. Simulation results are shown to illustrate the effectiveness of the proposed control schemes. [ABSTRACT FROM AUTHOR]

Many real-world classification problems are becoming multi-label in nature, i.e., multiple class labels are assigned to an instance simultaneously. Multi-label classification is a challenging problem due to the involvement of three forms of interactions, i.e., feature-to-feature, feature-to-label, and label-to-label interactions. What further complicates the problem is that not all features are useful, and some can deteriorate the classification performance. Sparsity-based methods have been widely used to address multi-label feature selection due to their efficiency and effectiveness. However, most (if not all) existing methods do not consider the three forms of interactions simultaneously, which could hinder their ability to achieve good performance. Moreover, most existing methods are gradient-based, which are prone to getting stuck at local optima. This paper proposes a new sparsity-based feature selection approach that can simultaneously consider all three forms of interactions. Furthermore, this paper develops a novel sparse learning method based on particle swarm optimisation that can avoid local optima. The proposed method is compared against the state-of-the-art multi-label feature selection methods in terms of multi-label classification performance. The results show that our method performed significantly better in selecting high-quality feature subsets with respect to various feature subset sizes. [ABSTRACT FROM AUTHOR]

Innovations in digital health and machine learning are changing the path of clinical health and care. People from different geographical locations and cultural backgrounds can benefit from the mobility of wearable devices and smartphones to monitor their health ubiquitously. This paper focuses on reviewing the digital health and machine learning technologies used in gestational diabetes – a subtype of diabetes that occurs during pregnancy. This paper reviews sensor technologies used in blood glucose monitoring devices, digital health innovations and machine learning models for gestational diabetes monitoring and management, in clinical and commercial settings, and discusses future directions. Despite one in six mothers having gestational diabetes, digital health applications were underdeveloped, especially the techniques that can be deployed in clinical practice. There is an urgent need to (1) develop clinically interpretable machine learning methods for patients with gestational diabetes, assisting health professionals with treatment, monitoring, and risk stratification before, during and after their pregnancies; (2) adapt and develop clinically-proven devices for patient self-management of health and well-being at home settings (“virtual ward” and virtual consultation), thereby improving clinical outcomes by facilitating timely intervention; and (3) ensure innovations are affordable and sustainable for all women with different socioeconomic backgrounds and clinical resources. [ABSTRACT FROM AUTHOR]

Human opacity is the intrinsic quality of unknowability of human beings with respect to machines. The descriptive relationship between humans and machines, which captures how much information one can gather about the other, can be explicated using an opacity-transparency relationship. This relationship allows us to describe and normatively evaluate a spectrum of opacity where humans and machines may be either opaque or transparent. In this paper, we argue that the advent of Affective Computing (AC) has begun to shift the ideal position of humans on this spectrum towards greater transparency, while much of this technology is shifting towards opacity. We explore the implications of this shift with regard to the affective information of humans and how the threat to human opacity by AC systems has various adverse repercussions, such as infringement of one's autonomy, deception, manipulation, and increased anxiety. There are also distributive consequences that expose vulnerable groups to unjustified burdens and reduce them to mere profiles. We further provide an assessment of current AC technology, which follows the descriptive relationship between humans and machines from the lens of opacity and transparency. Finally, we foresee and address three possible objections to our claims. These are the beneficence of AC systems, their relation to privacy, and their restrictive capacity to capture human affects. Through these arguments, the paper aims to bring attention to the ontological relationship between humans and machines from the perspective of opacity and transparency while emphasizing on the gravity of the ethical concerns raised by their threat to human opacity. [ABSTRACT FROM AUTHOR]

Reinforcement, extinction, generalization and differentiation are all basic principles of Pavlov associative memory. Most memristive neural networks that simulate associative memory only consider reinforcement and extinction, while ignoring differentiation and generalization. In this paper, a memristive circuit of associative memory with generalization and differentiation is proposed to solve the above problem. It implements the functions of learning, forgetting, long-term memory, generalization and differentiation. Learning and forgetting correspond to reinforcement and extinction in associative memory respectively. Spontaneous recovery, in which forgotten reflexes can reappear in the absence of an unconditional stimulus, is also discussed here. Besides, a special differentiation method that takes into account the time delay is designed and demonstrated. The proposed memristive circuit of associative memory provides a reference for the theoretical research and application of artificial neural networks. [ABSTRACT FROM AUTHOR]

Gasoline blending scheduling uses resource allocation and operation sequencing to meet a refinery's production requirements. The presence of nonlinearity, integer constraints, and a large number of decision variables adds complexity to this problem, posing challenges for traditional and evolutionary algorithms. This paper introduces a novel multiobjective optimization approach driven by a diffusion model (named DMO), which is designed specifically for gasoline blending scheduling. To address integer constraints and generate feasible schedules, the diffusion model creates multiple intermediate distributions between Gaussian noise and the feasible domain. Through iterative processes, the solutions transition from Gaussian noise to feasible schedules while optimizing the objectives using the gradient descent method. DMO achieves simultaneous objective optimization and constraint adherence. Comparative tests are conducted to evaluate DMO's performance across various scales. The experimental results demonstrate that DMO surpasses state-of-the-art multiobjective evolutionary algorithms in terms of efficiency when solving gasoline blending scheduling problems. [ABSTRACT FROM AUTHOR]

Evolutionary algorithms (EAs) have shown their competitiveness in solving the problem of feature selection. However, limited by their encoding scheme, most of them face the challenge of "curse of dimensionality". To address the issue, in this paper, an objective space constraint-based evolutionary algorithm, named OSC-EA, is proposed for high-dimensional feature selection (HDFS). Although the decision space of EAs for HDFS is very huge, its objective space is the same as that of the low-dimensional feature selection. Based on this fact, in the proposed OSC-EA, the HDFS is firstly modeled as a constrained problem, where a constraint of the objective space is introduced and used to partition the whole objective space into the "feasible region" and the "infeasible region". To handle the constrained problem, a two-stage $\varepsilon$ɛ constraint-based evolutionary scheme is designed. In the first stage, the value of $\varepsilon$ɛ is set to be very small, which ensures that the search concentrates on the "feasible region", and the latent high-quality feature subsets can be found quickly. Then, in the second stage, the value of $\varepsilon$ɛ increases gradually, so that more solutions in the "infeasible region" are considered. Until the end of the scheme, $\varepsilon \rightarrow \infty$ɛ→∞; all the solutions in the objective space are considered. By using the search in the second stage, the quality of the obtained feature subsets is further improved. The empirical results on different high-dimensional datasets demonstrate the effectiveness and efficiency of the proposed OSC-EA. [ABSTRACT FROM AUTHOR]

In this letter, a combined encryption and error-control coding scheme based on turbo codes (TCs) is proposed to enhance the security of the transmission. It relies on convolutional encoders that operate over Galois fields ${\textrm {GF}}(4)$ and include variable encryption polynomials, whose coefficients are randomly selected every $L$ steps from a set of optimal polynomials. This type of encoder has been studied in previous papers, where a certain correspondence between the randomness of the encoded sequence and the performance of the TC was conjectured. In this work, a scheme with a varying length $L$ is proposed to increase the output randomness and, therefore, the security of the sequence to be transmitted. That is to say, the number of steps $L$ , over which one of the polynomials is iterated, is a random variable. The randomness of the output and the security of the TC are quantified using differential block entropy, showing that it is possible to improve the security of the transmitted data without compromising the corresponding bit error rate (BER). [ABSTRACT FROM AUTHOR]

This paper provides an accessible explanation of the working mechanism of Monte-Carlo Tree Search, an influential search algorithm. The paper summarizes the procedure of Monte-Carlo Tree Search, including selection, expansion, simulation, and backpropagation. Additionally, immersive examples based on Tic-Tac-Toe, Go, and Sokoban, two two-player competitive games and a classic single-player puzzle game, are presented to illustrate how Monte-Carlo Tree Search works. The full article with interactive contents is published on IEEE Xplore. [ABSTRACT FROM AUTHOR]

In this paper, the idea of a fuzzy color and a fuzzy color space is shown in an interactive way. It is proposed several interactive elements, where readers can understand the different steps to build them. Furthermore, these elements allow the user to test with his/her own images via the behavior of the fuzzy colors. [ABSTRACT FROM AUTHOR]

Thanks to the many submissions we have received, we can present this second part of our special issue on "AI-eXplained." In here, we continue our mission to demystify the intricate world of artificial intelligence and make it accessible to a broader audience. As AI continues to evolve and integrate into various aspects of our lives, it becomes increasingly important to bridge the gap between experts and those eager to understand the inner workings of AI systems. Our goal remains unchanged—to present AI concepts in a comprehensible, engaging, and interactive manner, empowering our readers to explore and grasp the enchanting world of AI. [ABSTRACT FROM AUTHOR]

This paper proposes a linear regression model for scalar-valued responses and two-way functional (bivariate) predictors. Our motivation stems from the quality evaluation of products based on optical emission spectroscopy data from virtual metrology of semiconductor manufacturing. We focus on multivariate cases where the smoothness and shapes of the data vary significantly across variables. We propose a two-step solution to this problem, consisting of decomposition and prediction. First, we decompose the two-way functional data into pairs of component functions using functional singular value decomposition. Next, we build functional linear models for the decomposed functional variables and obtain the final predictor by averaging the models. Results from numerical studies, including simulation studies and real data analysis, demonstrate the promising empirical properties of the proposed approach, especially when the number of predictors is large. [ABSTRACT FROM AUTHOR]

It has been a widely growing interest in using silicon carbide (SiC) in high-power electronic devices. Yet, SiC wafers may contain killer defects that could reduce fabrication yield and make the device fall into unexpected failures. To prevent these failures from happening, it is very important to develop inspection tools that can detect, characterize and locate these defects in a non-invasive way. Current inspection techniques such as Dark Field or Bright field microscopy are effectively able to visualize most such defects; however, there are some scenarios where the inspection becomes problematic or almost impossible, such as when the defects are too small or have low contrast or if the defects lie deep into the substrate. Thus, an alternative method is needed to face these challenges. In this paper, we demonstrate the application of coherent Fourier scatterometry (CFS) as a complementary tool in addition to the conventional techniques to overcome different and problematic scenarios of killer defects inspection on SiC samples. Scanning electron microscopy (SEM) has been used to assess the same defects to validate the findings of CFS. Great consistency has been demonstrated in the comparison between the results obtained with CFS and SEM. [ABSTRACT FROM AUTHOR]

This paper addresses the quality control of the photolithography process in the semiconductor industry. Overlay errors in the process seriously affect the wafer yield, and cause the wafer to be forced to rework and affect the production efficiency of the equipment. We examine the current state of its process control, develop a novel overlay predict model, and verify the prediction results. This study proposes a Global Attention Generative Adversarial Networks (GAGAN) model to precisely predict the overlay error for the feed-forward data of the front layer, which is used as the important information and process parameters for the advanced process control of the current layer. Experiment results on a semiconductor shop-floor confirms that our proposed method achieves high predictive performance while maintaining extensibility and visual quality. [ABSTRACT FROM AUTHOR]

Recent developments in the field of non-local attention (NLA) have led to a renewed interest in self-similarity-based single image super-resolution (SISR). Researchers usually use the NLA to explore non-local self-similarity (NSS) in SISR and achieve satisfactory reconstruction results. However, a surprising phenomenon that the reconstruction performance of the standard NLA is similar to that of the NLA with randomly selected regions prompted us to revisit NLA. In this paper, we first analyzed the attention map of the standard NLA from different perspectives and discovered that the resulting probability distribution always has full support for every local feature, which implies a statistical waste of assigning values to irrelevant non-local features, especially for SISR which needs to model long-range dependence with a large number of redundant non-local features. Based on these findings, we introduced a concise yet effective soft thresholding operation to obtain high-similarity-pass attention (HSPA), which is beneficial for generating a more compact and interpretable distribution. Furthermore, we derived some key properties of the soft thresholding operation that enable training our HSPA in an end-to-end manner. The HSPA can be integrated into existing deep SISR models as an efficient general building block. In addition, to demonstrate the effectiveness of the HSPA, we constructed a deep high-similarity-pass attention network (HSPAN) by integrating a few HSPAs in a simple backbone. Extensive experimental results demonstrate that HSPAN outperforms state-of-the-art approaches on both quantitative and qualitative evaluations. Our code and a pre-trained model were uploaded to GitHub (https://github.com/laoyangui/HSPAN) for validation. [ABSTRACT FROM AUTHOR]

How to model the effect of reflection is crucial for single image reflection removal (SIRR) task. Modern SIRR methods usually simplify the reflection formulation with the assumption of linear combination of a transmission layer and a reflection layer. However, the large variations in image content and the real-world picture-taking conditions often result in far more complex reflection. In this paper, we introduce a new screen-blur combination based on two important factors, namely the intensity and the blurriness of reflection, to better characterize the reflection formulation in SIRR. Specifically, we present Screen-blur Reflection Networks (SRNet), which executes the screen-blur formulation in its network design and adapts to the complex reflection on real scenes. Technically, SRNet consists of three components: a blended image generator, a reflection estimator and a reflection removal module. The image generator exploits the screen-blur combination to synthesize the training blended images. The reflection estimator learns the reflection layer and a blur degree that measures the level of blurriness for reflection. The reflection removal module further uses the blended image, blur degree and reflection layer to filter out the transmission layer in a cascaded manner. Superior results on three different SIRR methods are reported when generating the training data on the principle of the screen-blur combination. Moreover, extensive experiments on six datasets quantitatively and qualitatively demonstrate the efficacy of SRNet over the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Efficient point cloud compression is essential for applications like virtual and mixed reality, autonomous driving, and cultural heritage. This paper proposes a deep learning-based inter-frame encoding scheme for dynamic point cloud geometry compression. We propose a lossy geometry compression scheme that predicts the latent representation of the current frame using the previous frame by employing a novel feature space inter-prediction network. The proposed network utilizes sparse convolutions with hierarchical multiscale 3D feature learning to encode the current frame using the previous frame. The proposed method introduces a novel predictor network for motion compensation in the feature domain to map the latent representation of the previous frame to the coordinates of the current frame to predict the current frame’s feature embedding. The framework transmits the residual of the predicted features and the actual features by compressing them using a learned probabilistic factorized entropy model. At the receiver, the decoder hierarchically reconstructs the current frame by progressively rescaling the feature embedding. The proposed framework is compared to the state-of-the-art Video-based Point Cloud Compression (V-PCC) and Geometry-based Point Cloud Compression (G-PCC) schemes standardized by the Moving Picture Experts Group (MPEG). The proposed method achieves more than 88% BD-Rate (Bjøntegaard Delta Rate) reduction against G-PCCv20 Octree, more than 56% BD-Rate savings against G-PCCv20 Trisoup, more than 62% BD-Rate reduction against V-PCC intra-frame encoding mode, and more than 52% BD-Rate savings against V-PCC P-frame-based inter-frame encoding mode using HEVC. These significant performance gains are cross-checked and verified in the MPEG working group. [ABSTRACT FROM AUTHOR]

Effectively evaluating the perceptual quality of dehazed images remains an under-explored research issue. In this paper, we propose a no-reference complex-valued convolutional neural network (CV-CNN) model to conduct automatic dehazed image quality evaluation. Specifically, a novel CV-CNN is employed that exploits the advantages of complex-valued representations, achieving better generalization capability on perceptual feature learning than real-valued ones. To learn more discriminative features to analyze the perceptual quality of dehazed images, we design a dual-stream CV-CNN architecture. The dual-stream model comprises a distortion-sensitive stream that operates on the dehazed RGB image, and a haze-aware stream on a novel dark channel difference image. The distortion-sensitive stream accounts for perceptual distortion artifacts, while the haze-aware stream addresses the possible presence of residual haze. Experimental results on three publicly available dehazed image quality assessment (DQA) databases demonstrate the effectiveness and generalization of our proposed CV-CNN DQA model as compared to state-of-the-art no-reference image quality assessment algorithms. [ABSTRACT FROM AUTHOR]

Small moving objects at far distance always occupy only one or a few pixels in image and exhibit extremely limited visual features, which bring great challenges to motion detection. Highly evolved visual systems endow flying insects with remarkable ability to pursue tiny mates and prey, providing a good template to develop image processing method for small target motion detection. The insects’ excellent sensitivity to small moving objects is believed to come from a class of specific neurons called small target motion detectors (STMDs). However, existing STMD-based methods often experience performance degradation when coping with complex natural scenes. In this paper, we propose a bio-inspired visual system with spatio-temporal feedback mechanism (called Spatio-Temporal Feedback STMD) to suppress false positive background movement while enhancing system responses to small targets. Specifically, the proposed visual system is composed of two complementary subnetworks and a feedback loop. The first subnetwork is designed to extract spatial and temporal movement patterns of cluttered background by neuronal ensemble coding. The second subnetwork is developed to capture small target motion information where its output and signal from the first subnetwork are integrated together via the feedback loop to filter out background false positives in a recurrent manner. Experimental results demonstrate that the proposed spatio-temporal feedback visual system is more competitive than existing methods in discriminating small moving targets from complex natural environments. [ABSTRACT FROM AUTHOR]

Unmanned Aerial Vehicles (UAVs) rely on satellite systems for stable positioning. However, due to limited satellite coverage or communication disruptions, UAVs may lose signals for positioning. In such situations, vision-based techniques can serve as an alternative, ensuring the self-positioning capability of UAVs. However, most of the existing datasets are developed for the geo-localization task of the objects captured by UAVs, rather than UAV self-positioning. Furthermore, the existing UAV datasets apply discrete sampling to synthetic data, such as Google Maps, neglecting the crucial aspects of dense sampling and the uncertainties commonly experienced in practical scenarios. To address these issues, this paper presents a new dataset, DenseUAV, that is the first publicly available dataset tailored for the UAV self-positioning task. DenseUAV adopts dense sampling on UAV images obtained in low-altitude urban areas. In total, over 27K UAV- and satellite-view images of 14 university campuses are collected and annotated. In terms of methodology, we first verify the superiority of Transformers over CNNs for the proposed task. Then we incorporate metric learning into representation learning to enhance the model’s discriminative capacity and to reduce the modality discrepancy. Besides, to facilitate joint learning from both the satellite and UAV views, we introduce a mutually supervised learning approach. Last, we enhance the Recall@K metric and introduce a new measurement, SDM@K, to evaluate both the retrieval and localization performance for the proposed task. As a result, the proposed baseline method achieves a remarkable Recall@1 score of 83.01% and an SDM@1 score of 86.50% on DenseUAV. The dataset and code have been made publicly available on https://github.com/Dmmm1997/DenseUAV. [ABSTRACT FROM AUTHOR]

This paper focuses on the facial micro-expression (FME) generation task, which has potential application in enlarging digital FME datasets, thereby alleviating the lack of training data with labels in existing micro-expression datasets. Despite obvious progress in the image animation task, FME generation remains challenging because existing image animation methods can hardly encode subtle and short-term facial motion information. To this end, we present a facial-prior-guided FME generation framework that takes advantage of facial priors for facial motion generation. Specifically, we first estimate the geometric locations of action units (AUs) with detected facial landmarks. We further calculate an adaptive weighted prior (AWP) map, which alleviates the estimation error of AUs while efficiently capturing subtle facial motion patterns. To achieve smooth and realistic synthesis results, we use our proposed facial prior module to guide motion representation and generation modules in mainstream image animation frameworks. Extensive experiments on three benchmark datasets consistently show that our proposed facial prior module can be adopted in image animation frameworks and significantly improve their performance on micro-expression generation. Moreover, we use the generation technique to enlarge existing datasets, thereby improving the performance of general action recognition backbones on the FME recognition task. Our code is available at https://github.com/sysu19351158/FPB-FOMM. [ABSTRACT FROM AUTHOR]

Most researchers focus on designing accurate crowd counting models with heavy parameters and computations but ignore the resource burden during the model deployment. A real-world scenario demands an efficient counting model with low-latency and high-performance. Knowledge distillation provides an elegant way to transfer knowledge from a complicated teacher model to a compact student model while maintaining accuracy. However, the student model receives the wrong guidance with the supervision of the teacher model due to the inaccurate information understood by the teacher in some cases. In this paper, we propose a dual-knowledge distillation (DKD) framework, which aims to reduce the side effects of the teacher model and transfer hierarchical knowledge to obtain a more efficient counting model. First, the student model is initialized with global information transferred by the teacher model via adaptive perspectives. Then, the self-knowledge distillation forces the student model to learn the knowledge by itself, based on intermediate feature maps and target map. Specifically, the optimal transport distance is utilized to measure the difference of feature maps between the teacher and the student to perform the distribution alignment of the counting area. Extensive experiments are conducted on four challenging datasets, demonstrating the superiority of DKD. When there are only approximately 6% of the parameters and computations from the original models, the student model achieves a faster and more accurate counting performance as the teacher model even surpasses it. [ABSTRACT FROM AUTHOR]

Covert communication is an method that plays an important role in secure data transmission. The technology embeds covert information into data and propagates it through covert channels. The communication quality depends on the choice of channel and data embedding techniques. Recently, blockchain has emerged to become the preferred channel to carry out covert communication for its decentralization and anonymity features. Existing covert transaction methods are constructed transaction-by-transaction, which makes them immune to text analysis-based detection methods. However, it is easy to expose their features on the transaction graph level. Unfortunately, there is yet no method to detect covert transactions by the features of transaction graph. In this paper, we propose a covert transaction detection method based on graph structure. By analyzing the statistical features of graph structure for addresses, we can infer whether they are the participants of covert transactions. Furthermore, we design a protection method of covert transactions based on graph generation networks. By adjusting the structural features between different addresses, our method enhances the security of multiple interrelated covert transactions. Experimental analysis on the Bitcoin Testnet verifies the security and the efficiency of the proposed methods. [ABSTRACT FROM AUTHOR]

E-healthcare as a significant facet of the Internet of Things (IoT) relies on wearable devices to continuously monitor users’ vital signals for health-related purposes. The sensitive and vast nature of the collected data necessitate secure encryption and storage on the cloud. Simultaneously, there is a need to share these data for healthcare purposes. How to balance the privacy and usability of these data is a challenging problem in the e-healthcare applications. A central problem arose from the continuous data collection is how to effectively share the data collected in one specific time period when users are unwell. We formalize it as the sequential data sharing problem. This problem appears in various IoT applications apart from e-healthcare, such as video surveillance. In this paper, we explain why all existing solutions cannot address the above problem. Then, we propose a novel sequential data sharing system (SDSS), where the data encrypted at any specific time period can be efficiently shared. We first present a practical construction of SDSS that supports securely sharing data within one specific time period in a symmetric manner. The decryption key are retrieved sequentially by utilizing the shared key and hash function. All involved calculations in the system are lightweight. Data pertaining to other non-selected time periods remain unknown. We then extend the SDSS to a multiple-range version, enabling simultaneous sharing data for multiple specific time periods, and show an example. We formalize the definition of security models and analyze the security of our systems. Finally, we evaluate the performance of our systems using SHA-256 and AES-256. Experimental results demonstrate that our systems are highly efficient. It takes less than 1 millisecond to encrypt 100KB data and less than 0.4 milliseconds to decrypt the corresponding cipher data. Our proposed systems provide a solution to balance the privacy and usability in the context of sequentially collected data. We believe that this work will enhance the adoption and practicality of IoT applications. [ABSTRACT FROM AUTHOR]

With the growing use of mobile devices, location-based services (LBS) are becoming increasingly popular. BLS deliver accurate services to individuals according to their geographical locations, but privacy issues have been the primary concerns of users. Privacy-preserving LBS (PPLBS) were proposed to protect location privacy, but there are still some problems: 1) a semi-trusted third party (STTP) is required to blur users’ locations; 2) both the computation and communication costs of generating a query are linear with the size of queried areas; 3) the schemes were not formally treated, in terms of definition, security model, security proof, etc. In this paper, to protect location privacy and improve query efficiency, an oblivious location-based services (OLBS) scheme is proposed. Our scheme captures the following features: 1) an STTP is not required; 2) users can query services without revealing their exact location information; 3) the service provider only knows the size of queried areas and nothing else; and 4) both the computation and communication costs of generating a query is constant, instead of linear with the size of queried areas. We formalise both the definition and security model of our OLBS scheme, and propose a concrete construction. Furthermore, the implementation is conducted to show its efficiency. The security of our scheme is reduced to well-known complexity assumptions. The novelty is to reduce the computation and communication costs of generating a query and enable the service provider to obliviously generate decrypt keys for queried services. This contributes to the growing work of formalising PPLBS schemes and improving query efficiency. [ABSTRACT FROM AUTHOR]

Heterogeneous pedestrian images are captured by visible and infrared cameras with different spectrums, which play an important role in night-time video surveillance. However, visible infrared person re-identification (VI-REID) is still a challenging problem due to the considerable cross-modality discrepancies. To extract modality-invariant features which are discriminative for the person identity, recent studies are inclined to regard modality-specific features as noise and discard them. Actually, the modality-specific characteristics containing background and color information are indispensable for learning modality-shared features. In this paper, we propose a novel Dual-Adversarial Representation Disentanglement (DARD) model to separate modality-specific features from tangled pedestrian representations and effectively learn the robust modality-invariant representations. Specifically, our method employs dual-adversarial learning, incorporating image-level channel exchange and feature-level magnitude change to introduce variations in modality-specific representations. This deliberate perturbation raises the learning difficulty for the model to learn modality-shared features. Simultaneously, to control the changing scope of modality-specific features, bi-constrained noise alleviation is introduced during adversarial learning, keeping the balance of feature generation and adversary. The proposed dual-adversarial learning methodology enhances the robustness against cross-modality visual discrepancy and strengthens the discriminative power of the learned modality-shared representations without introducing additional network parameters. This improvement further elevates the retrieval performance of VI-REID. Extensive experiments with insightful analysis on two cross-modality re-identification datasets verify the effectiveness and superiority of the proposed DARD method. [ABSTRACT FROM AUTHOR]