Research Directions Research Articles

A Survey of Visual Transformers.

Transformer, an attention-based encoder-decoder model, has already revolutionized the field of natural language processing (NLP). Inspired by such significant achievements, some pioneering works have recently been done on employing Transformer-liked architectures in the computer vision (CV) field, which have demonstrated their effectiveness on three fundamental CV tasks (classification, detection, and segmentation) as well as multiple sensory data stream (images, point clouds, and vision-language data). Because of their competitive modeling capabilities, the visual Transformers have achieved impressive performance improvements over multiple benchmarks as compared with modern convolution neural networks (CNNs). In this survey, we have reviewed over 100 of different visual Transformers comprehensively according to three fundamental CV tasks and different data stream types, where taxonomy is proposed to organize the representative methods according to their motivations, structures, and application scenarios. Because of their differences on training settings and dedicated vision tasks, we have also evaluated and compared all these existing visual Transformers under different configurations. Furthermore, we have revealed a series of essential but unexploited aspects that may empower such visual Transformers to stand out from numerous architectures, e.g., slack high-level semantic embeddings to bridge the gap between the visual Transformers and the sequential ones. Finally, two promising research directions are suggested for future investment. We will continue to update the latest articles and their released source codes at https://github.com/liuyang-ict/awesome-visual-transformers.

MAD-Former: A Traceable Interpretability Model for Alzheimer's Disease Recognition Based on Multi-Patch Attention.

The integration of structural magnetic resonance imaging (sMRI) and deep learning techniques is one of the important research directions for the automatic diagnosis of Alzheimer's disease (AD). Despite the satisfactory performance achieved by existing voxel-based models based on convolutional neural networks (CNNs), such models only handle AD-related brain atrophy at a single spatial scale and lack spatial localization of abnormal brain regions based on model interpretability. To address the above limitations, we propose a traceable interpretability model for AD recognition based on multi-patch attention (MAD-Former). MAD-Former consists of two parts: recognition and interpretability. In the recognition part, we design a 3D brain feature extraction network to extract local features, followed by constructing a dual-branch attention structure with different patch sizes to achieve global feature extraction, forming a multi-scale spatial feature extraction framework. Meanwhile, we propose an important attention similarity position loss function to assist in model decision-making. The interpretability part proposes a traceable method that can obtain a 3D ROI space through attention-based selection and receptive field tracing. This space encompasses key brain tissues that influence model decisions. Experimental results reveal the significant role of brain tissues such as the Fusiform Gyrus (FuG) in AD recognition. MAD-Former achieves outstanding performance in different tasks on ADNI and OASIS datasets, demonstrating reliable model interpretability.

Split Learning with Differential Privacy for Integrated Terrestrial and Non-Terrestrial Networks

Integrated terrestrial and non-terrestrial networks (TNTNs) have become promising architecture for enabling ubiquitous connectivity. Smart remote sensing is one of the typical applications of TNTNs that collects and analyzes various dimensions of remote sensing data by deploying Internet of Things (IoT) sensors and edge computing in terrestrial, space, aerial, and underwater networks. To improve the analysis accuracy of remote sensing data, the owners of different networks should conduct collaborative learning on different dimensions of data, while data and label privacy should be jointly considered. However, the existing collaborative learning paradigms are difficult to meet this demand. In this article, we propose a paradigm of split learning (SL), where the data owner and the label owner train different parts of the deep learning model and only exchange the intermediate data (i.e., smashed data and cut layer gradients). We explore the potential privacy attacks that recover the raw data and label information based on the intermediate data and propose the differential privacy (DP) based defense mechanisms that inject randomly generated Laplace noise into the intermediate data to disturb the attack performance. We also conduct a simulation study based on a real-world satellite remote sensing dataset to validate that the SL paradigm with defense mechanisms can effectively balance the performance of the collaborative model training and the protection of the data and label privacy. Finally, we discuss the main challenges and potential research directions of the privacy-preserving SL paradigm for smart remote sensing over TNTNs.

Practitioner Perspectives on Going Concern Opinion Research and Suggestions for Further Study: Part 2—Reporting Environment and the Decision-Making Process

SYNOPSIS Auditing researchers rarely engage practicing auditors regarding the phenomena examined, or conclusions reached, in academic research. In an attempt to bridge this gap, we report on the outcomes, insights, and observations from focused interactions with Dutch audit practitioners regarding their perspectives on recent research regarding auditor going concern opinion (GCO) decisions. We provided practitioners with a summary of some of the most salient recent GCO research findings. The goals of our subsequent discussions were to (1) obtain practitioners’ responses to the research findings, and (2) identify relevant issues for future study from the perspective of practitioners. Accordingly, in this second of a two-part sequence of articles, we discuss the outcomes of our interactions with practitioners related to the GCO reporting environment and the GCO decision-making process, as well as directions for future research. JEL Classifications: M42.

Recent advances in the study of epitopes, allergens and immunologic cross-reactivity of edible mango

Mango (Mangifera indica L.) is a tropical fruit that is widely consumed as both fresh fruits and processed products around the world. The high incidence of mango allergy, on the other hand, has sparked widespread concern. Therefore, a summary and analysis of the current status and issues in mango allergen research can guide in-depth study on the mechanism of mango allergy and reveal effective desensitization methods. We described the incidence of fruit allergy, as well as the mechanism and clinical symptoms of mango allergy, in this review. We also looked into the structural properties of mango allergens, the effect of processing methods on mango allergens, prediction methods for mango allergen epitopes, and the current state of research on mango cross-reactive allergens and preventive measures. Finally, the research directions and ideas for the future are proposed and discussed.

Switchable Signaling Molecules for Media Modulation: Fundamentals, Applications, and Research Directions

Switchable Signaling Molecules for Media Modulation: Fundamentals, Applications, and Research Directions

Taking a heavier toll? Racial differences in the effects of workplace mistreatment on depression.

Previous studies have found that workplace mistreatment positively relates to depression, a critical mental health disorder. However, it is unknown whether mistreatment affects all individuals' depressive symptoms equally. Drawing from the hopelessness theory of depression and the stigma literature, we suggest that Blacks suffer from greater depression than Whites when they experience similar levels of workplace mistreatment because Blacks, as members of a racial minority group, are more likely to attribute workplace mistreatment to their race. This, in turn, causes them to make a pessimistic attribution (i.e., attributions that are internal, stable, and global) about themselves that, ultimately, leads to depression. We tested these predictions across two studies. In Study 1, we used a multiyear time-lagged design and multiple indicators of depression (i.e., self-reported clinical depression scale, device-traced sleep quantity, and self-reported sleep quality) and found that the positive relationship between workplace mistreatment and depression was stronger for Blacks than Whites, and that these patterns were consistent across the various indicators (although only results with the clinical depression scale and sleep quantity were statistically significant). In Study 2, we found that the influence of workplace mistreatment on depression is partly due to racial differences in how workplace mistreatment is attributed. We conclude by discussing the theoretical and practical implications of these findings and directions for future research. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Muscle Steatosis and Fibrosis in Older Adults, From the AJR Special Series on Imaging of Fibrosis.

The purpose of this article is to review steatosis and fibrosis of skeletal muscle, focusing on older adults. Although CT, MRI, and ultrasound are commonly used to image skeletal muscle and provide diagnoses for a variety of medical conditions, quantitative assessment of muscle steatosis and fibrosis is uncommon. This review provides radiologists with a broad perspective on muscle steatosis and fibrosis in older adults by considering the public health impact, biologic mechanisms, and evaluation with CT, MRI, and ultrasound. Promising directions in clinical research that entail artificial intelligence algorithms and the imaging assessment of biologic age are also reviewed. The imaging methods presented hold promise for improving the evaluation of common conditions affecting older adults, including sarcopenia, frailty, and cachexia.

Targeting NLRP3 Inflammasome: Structure, Function, and Inhibitors.

Inflammasomes are multimeric protein complexes that can detect various physiological stimuli and danger signals. As a result, they perform a crucial function in the innate immune response. The NLRP3 inflammasome, as a vital constituent of the inflammasome family, is significant in defending against pathogen invasion and preserving cellhomeostasis. NLRP3 inflammasome dysregulation is connected to various pathological conditions, including inflammatory diseases, cancer, and cardiovascular and neurodegenerative diseases. This profile makes NLRP3 an applicable target for treating related diseases, and therefore, there are rising NLRP3 inhibitors disclosed for therapy. Herein, we summarized the updated advances in the structure, function, and inhibitors of NLRP3 inflammasome. Moreover, we aimed to provide an overview of the existing products and future directions for drug research and development.

Review of Patents Analysis on Floor Robot Vacuum

Background: As the pace of life gets faster, and with the development and progress of society and the continuous improvement of people's pursuit of a convenient and comfortable life, it has become a growing desire for people to get rid of tedious housework. Hygienic automatic cleaning has also become the main demand of people. Objective: Through review, analysis, and discussion of the patents on existing floor robot vacuums, it is expected to find out the development process and main outcomes, along with the main technologies involved, and provide a valuable prediction basis for future research and development directions. Methods: This study aimed to summarize a number of representative patents related to floor robot vacuums and their main components and analyze them in depth to study their technical points and development laws. Results: By summarizing a large number of patents on floor robot vacuums and further analyzing and comparing the characteristics of the patents, the future development trend of floor robot vacuums was further discussed. Conclusion: Through a detailed analysis of the mechanical structure, intelligent control technology, and sensor technology of the existing floor robot vacuums as a whole and the main work unit, this study discussed the ideas and methods of improving the new floor robot vacuum, which is of great significance to its operation and more convenient and intelligent.

Review of Recent Patents on Smart Bearing

Background: Smart bearing is based on the traditional bearing, which is integrated by sensing devices and control devices, formating a unique bearing structure unit. It uses information processing, automatic control, and other technologies to achieve real-time online monitoring of bearing operating conditions, fault detection, and state control. Compared with ordinary bearings, smart bearings have the characteristics of self-awareness, self-decision, and self-regulation. By analyzing the composition of various types of smart bearings and their working principles, it is beneficial to prospect the future development direction of bearings. Objective: We analyze and organize the patents on smart bearings in recent years, summarize the problems encountered in the development of smart bearings, and propose the development direction of smart bearings in the future to provide a reference for future scientific research. Methods: This paper reviews the patents related to smart bearings and analyzes the composition and structure of smart bearings. The main smart bearings include rolling bearing, sliding bearing, magnetic bearing, and rubber bearing. The principle and structure of these bearings are analyzed. Their advantages and shortcomings are discussed. Results: Through the analysis of relevant patents and papers on smart bearings, the current development status and problems of smart bearings are summarized, the future design of smart bearings is proposed, and the research ideas and directions are proposed. Conclusion: The development of smart bearings is conducive to the timely detection of bearing conditions in the operation process. Except for detection, smart bearings have the functions of judgment, feedback, processing, etc. There are many factors that interfere with the accuracy of the smart bearing judgment and information transfer, so it is necessary to improve the structure and working principle of smart bearings through a variety of measures. In the future, more smart bearings will be invented.

Toward a Framework for Trust Building between Humans and Robots in the Construction Industry: A Systematic Review of Current Research and Future Directions

Toward a Framework for Trust Building between Humans and Robots in the Construction Industry: A Systematic Review of Current Research and Future Directions

Recent Advances in Molecular Imprinting for Proteins on Magnetic Microspheres.

The separation of proteins in biological samples plays an essential role in the development of disease detection, drug discovery, and biological analysis. Protein imprinted polymers (PIPs) serve as a tool to capture target proteins specifically and selectively from complex media for separation purposes. Whereas conventional molecularly imprinted polymer is time-consuming in terms of incubation studies and solvent removal, magnetic particles are introduced using their magnetic properties for sedimentation and separation, resulting in saving extraction and centrifugation steps. Magnetic protein imprinted polymers (MPIPs), which combine molecularly imprinting materials with magnetic properties, have emerged as a new area of research hotspot. This review provides an overview of MPIPs for proteins, including synthesis, preparation strategies, and applications. Moreover, it also looks forward to the future directions for research in this emerging field.

Environmental Risk Factors in Autism Spectrum Disorder: A Narrative Review.

Existing evidence indicates that environmental factors might contribute up to 50% of the variance in autism spectrum disorder (ASD) risk. This structured narrative review offers a comprehensive synthesis of current knowledge on environmental risk factors in ASD, including evaluation of conflicting evidence, exploration of underlying mechanisms, and suggestions for future research directions. Analysis of diverse epidemiological investigations indicates that certain environmental factors, including advanced parental age, preterm birth, delivery complications, and exposure to toxic metals, drugs, air pollutants, and endocrine-disrupting chemicals, are linked to an increased ASD risk through various mechanisms such as oxidative stress, inflammation, hypoxia, and its consequences, changes in neurotransmitters, disruption of signaling pathways and some others. On the other hand, pregnancy-related factors such as maternal diabetes, maternal obesity, and caesarian section show a weaker association with ASD risk. At the same time, other environmental factors, such as vaccination, maternal smoking, or alcohol consumption, are not linked to the risk of ASD. Regarding nutritional elements data are inconclusive. These findings highlight the significance of environmental factors in ASD etiology and emphasize that more focused research is needed to target the risk factors of ASD. Environmental interventions targeting modifiable risk factors might offer promising avenues for ASD prevention and treatment.

IM-IAD: Industrial Image Anomaly Detection Benchmark in Manufacturing.

Image anomaly detection (IAD) is an emerging and vital computer vision task in industrial manufacturing (IM). Recently, many advanced algorithms have been reported, but their performance deviates considerably with various IM settings. We realize that the lack of a uniform IM benchmark is hindering the development and usage of IAD methods in real-world applications. In addition, it is difficult for researchers to analyze IAD algorithms without a uniform benchmark. To solve this problem, we propose a uniform IM benchmark, for the first time, to assess how well these algorithms perform, which includes various levels of supervision (unsupervised versus fully supervised), learning paradigms (few-shot, continual and noisy label), and efficiency (memory usage and inference speed). Then, we construct a comprehensive IAD benchmark (IM-IAD), which includes 19 algorithms on seven major datasets with a uniform setting. Extensive experiments (17 017 total) on IM-IAD provide in-depth insights into IAD algorithm redesign or selection. Moreover, the proposed IM-IAD benchmark challenges existing algorithms and suggests future research directions. For reproducibility and accessibility, the source code is uploaded to the website: https://github.com/M-3LAB/open-iad.

The Metaverse for Cognitive Health: A Paradigm Shift

The Metaverse, frequently dubbed as ”the successor of mobile internet”, is a paradigm which aims to build an immersive, spatiotemporal and self-sustained virtual, but realistic world for all age groups. The Metaverse is stepping from science fiction to reality due to the advances in technologies like Extended Reality, Virtual Reality, Augmented Reality, Artificial Intelligence, the Internet of Everything, eXtended meta-uni-omni-Verse and Digital Twins. Consumer electronics such as wearables, television, and smartphones are essential components of the Metaverse that can be used to accurately collect and measure data concerning user interaction and thereby create an immersive environment. This article discusses the state of art literature about how the Metaverse can be used for improving cognitive health, various scenarios where the Metaverse could help doctors to better treat cognitive patients, and finally, the future research directions that can realize the potential of the Metaverse for cognitive health.

A Survey on Deep Learning Event Extraction: Approaches and Applications.

Event extraction (EE) is a crucial research task for promptly apprehending event information from massive textual data. With the rapid development of deep learning, EE based on deep learning technology has become a research hotspot. Numerous methods, datasets, and evaluation metrics have been proposed in the literature, raising the need for a comprehensive and updated survey. This article fills the research gap by reviewing the state-of-the-art approaches, especially focusing on the general domain EE based on deep learning models. We introduce a new literature classification of current general domain EE research according to the task definition. Afterward, we summarize the paradigm and models of EE approaches, and then discuss each of them in detail. As an important aspect, we summarize the benchmarks that support tests of predictions and evaluation metrics. A comprehensive comparison among different approaches is also provided in this survey. Finally, we conclude by summarizing future research directions facing the research area.

Hierarchical Design of Tissue-Mimetic Fibrillar Hydrogel Scaffolds.

Most tissues of the human body present hierarchical fibrillar extracellular matrices (ECMs) that have a strong influence over their physicochemical properties and biological behavior. Of great interest is the introduction of this fibrillar structure to hydrogels, particularly due to the water-rich composition, cytocompatibility, and tunable properties of this class of biomaterials. Here, the main bottom-up fabrication strategies for the design and production of hierarchical biomimetic fibrillar hydrogels and their most representative applications in the fields of tissue engineering and regenerative medicine are reviewed. For example, the controlled assembly/arrangement of peptides, polymeric micelles, cellulose nanoparticles (NPs), and magnetically responsive nanostructures, among others, into fibrillar hydrogels is discussed, as well as their potential use as fibrillar-like hydrogels (e.g., those from cellulose NPs) with key biofunctionalities such as electrical conductivity or remote stimulation. Finally, the major remaining barriers to the clinical translation of fibrillar hydrogels and potential future directions of research in this field are discussed.

Nominal-phrase premodifications by non-native English speakers: The case of attributive adjective word order of Saudi EFL students at tertiary level

English nominal premodifications such as adjective word order have piqued the interest of researchers in recent years due to their learning difficulty for all language learners regardless of their linguistic backgrounds. Non-native English speakers like EFL students may find this case challenging. In light of this, the current study sought to ascertain the difficulty level of Saudi EFL students in recognising the natural order of English attributive adjectives, taking into account the most challenging order sequence, gender, and language proficiency levels. To accomplish this main objective, a designed pronominal adjective test was developed and carried out on 139 Saudi undergraduate EFL students who were purposively recruited for this study. According to the findings, Saudi EFL students had varying degrees of difficulty recognising the natural English adjective word order. While 66.2% of students had moderate difficulty choosing the appropriate adjective sequence, only 15.1% and 18.7% of participants had low and high difficulty, respectively. The most difficult was nominal premodifications with four modifiers, followed by three and two modifiers. When there were only two modifiers, most participants correctly identified incorrect adjective word ordering and provided corrections. When more modifiers existed, the percentages decreased, and the difficulty level increased. Proficiency language level and gender statistically significant differences were also discovered. Although the former revealed no differences, the latter had differences in favour of males, implying that females performed better in the test than males. Based on these findings, limitations and future research directions were proposed.

Fuzzers for Stateful Systems: Survey and Research Directions

Fuzzing is a very effective testing methodology to find bugs. In a nutshell, a fuzzer sends many slightly malformed messages to the software under test, hoping for crashes or incorrect system behaviour. The methodology is relatively simple, although applications that keep internal states are challenging to fuzz. The research community has responded to this challenge by developing fuzzers tailored to stateful systems, but a clear understanding of the variety of strategies is still missing. In this paper, we present the first taxonomy of fuzzers for stateful systems and provide a systematic comparison and classification of these fuzzers.