Showing papers by "Nanyang Technological University published in 2021"
University of Southern California1, French Institute for Research in Computer Science and Automation2, University of Oulu3, Princeton University4, University of Warwick5, Georgia Institute of Technology6, Rutgers University7, University of Virginia8, University of Washington9, Carnegie Mellon University10, École Polytechnique Fédérale de Lausanne11, University of Pittsburgh12, University of Wisconsin-Madison13, University of California, San Diego14, University of Illinois at Urbana–Champaign15, Nanyang Technological University16, Australian National University17, Stanford University18, IT University of Copenhagen19, Massachusetts Institute of Technology20, University of California, Berkeley21, Cornell University22, Emory University23, Hong Kong University of Science and Technology24
TL;DR: In this article, the authors describe the state-of-the-art in the field of federated learning from the perspective of distributed optimization, cryptography, security, differential privacy, fairness, compressed sensing, systems, information theory, and statistics.
Abstract: The term Federated Learning was coined as recently as 2016 to describe a machine learning setting where multiple entities collaborate in solving a machine learning problem, under the coordination of a central server or service provider. Each client’s raw data is stored locally and not exchanged or transferred; instead, focused updates intended for immediate aggregation are used to achieve the learning objective. Since then, the topic has gathered much interest across many different disciplines and the realization that solving many of these interdisciplinary problems likely requires not just machine learning but techniques from distributed optimization, cryptography, security, differential privacy, fairness, compressed sensing, systems, information theory, statistics, and more. This monograph has contributions from leading experts across the disciplines, who describe the latest state-of-the art from their perspective. These contributions have been carefully curated into a comprehensive treatment that enables the reader to understand the work that has been done and get pointers to where effort is required to solve many of the problems before Federated Learning can become a reality in practical systems. Researchers working in the area of distributed systems will find this monograph an enlightening read that may inspire them to work on the many challenging issues that are outlined. This monograph will get the reader up to speed quickly and easily on what is likely to become an increasingly important topic: Federated Learning.
2,144 citations
TL;DR: A comprehensive review of the knowledge graph covering overall research topics about: 1) knowledge graph representation learning; 2) knowledge acquisition and completion; 3) temporal knowledge graph; and 4) knowledge-aware applications and summarize recent breakthroughs and perspective directions to facilitate future research.
Abstract: Human knowledge provides a formal understanding of the world. Knowledge graphs that represent structural relations between entities have become an increasingly popular research direction toward cognition and human-level intelligence. In this survey, we provide a comprehensive review of the knowledge graph covering overall research topics about: 1) knowledge graph representation learning; 2) knowledge acquisition and completion; 3) temporal knowledge graph; and 4) knowledge-aware applications and summarize recent breakthroughs and perspective directions to facilitate future research. We propose a full-view categorization and new taxonomies on these topics. Knowledge graph embedding is organized from four aspects of representation space, scoring function, encoding models, and auxiliary information. For knowledge acquisition, especially knowledge graph completion, embedding methods, path inference, and logical rule reasoning are reviewed. We further explore several emerging topics, including metarelational learning, commonsense reasoning, and temporal knowledge graphs. To facilitate future research on knowledge graphs, we also provide a curated collection of data sets and open-source libraries on different tasks. In the end, we have a thorough outlook on several promising research directions.
1,025 citations
TL;DR: A review on interpretabilities suggested by different research works and categorize them is provided, hoping that insight into interpretability will be born with more considerations for medical practices and initiatives to push forward data-based, mathematically grounded, and technically grounded medical education are encouraged.
Abstract: Recently, artificial intelligence and machine learning in general have demonstrated remarkable performances in many tasks, from image processing to natural language processing, especially with the advent of deep learning (DL). Along with research progress, they have encroached upon many different fields and disciplines. Some of them require high level of accountability and thus transparency, for example, the medical sector. Explanations for machine decisions and predictions are thus needed to justify their reliability. This requires greater interpretability, which often means we need to understand the mechanism underlying the algorithms. Unfortunately, the blackbox nature of the DL is still unresolved, and many machine decisions are still poorly understood. We provide a review on interpretabilities suggested by different research works and categorize them. The different categories show different dimensions in interpretability research, from approaches that provide “obviously” interpretable information to the studies of complex patterns. By applying the same categorization to interpretability in medical research, it is hoped that: 1) clinicians and practitioners can subsequently approach these methods with caution; 2) insight into interpretability will be born with more considerations for medical practices; and 3) initiatives to push forward data-based, mathematically grounded, and technically grounded medical education are encouraged.
810 citations
Cornell University1, National University of Singapore2, University of New South Wales3, University of Lausanne4, University of Michigan5, Erasmus University Rotterdam6, Tel Aviv University7, University of Melbourne8, Singapore Management University9, University of Maryland, College Park10, University of Pennsylvania11, Eindhoven University of Technology12, Stanford University13, Concordia University14, London Business School15, Baylor University16, University College London17, California State University, Sacramento18, INSEAD19, Saint Louis University20, Nanyang Technological University21, University of Minnesota22, Harvard University23, University of Arkansas24, VU University Amsterdam25
TL;DR: A broad-scope overview provides an integrative approach for considering the implications of COVID-19 for work, workers, and organizations while also identifying issues for future research and insights to inform solutions.
Abstract: The impacts of COVID-19 on workers and workplaces across the globe have been dramatic. This broad review of prior research rooted in work and organizational psychology, and related fields, is intended to make sense of the implications for employees, teams, and work organizations. This review and preview of relevant literatures focuses on (a) emergent changes in work practices (e.g., working from home, virtual teamwork) and (b) emergent changes for workers (e.g., social distancing, stress, and unemployment). In addition, potential moderating factors (demographic characteristics, individual differences, and organizational norms) are examined given the likelihood that COVID-19 will generate disparate effects. This broad-scope overview provides an integrative approach for considering the implications of COVID-19 for work, workers, and organizations while also identifying issues for future research and insights to inform solutions. (PsycInfo Database Record (c) 2020 APA, all rights reserved).
654 citations
Ludwig Maximilian University of Munich1, University of Cambridge2, University of Hyderabad3, Katholieke Universiteit Leuven4, Massachusetts Institute of Technology5, University of Antwerp6, Indian Institute of Science7, Yantai University8, King Abdullah University of Science and Technology9, University of Vigo10, Nanjing University of Science and Technology11, Lawrence Berkeley National Laboratory12, University of California, Berkeley13, Nanyang Technological University14, Soochow University (Suzhou)15, Technische Universität München16, ETH Zurich17, Lund University18, Hokkaido University19, University of California, Santa Cruz20, Chinese Academy of Sciences21, Beijing Institute of Technology22, City University of Hong Kong23, University of Texas at Austin24, Indian Association for the Cultivation of Science25, San Diego State University26, University of Washington27, Texas A&M University28, Bilkent University29, James I University30, Max Planck Society31, National Renewable Energy Laboratory32, University of Valencia33, Shanghai Jiao Tong University34, Istituto Italiano di Tecnologia35, Swiss Federal Laboratories for Materials Science and Technology36, University of Notre Dame37, Monash University, Clayton campus38, Imperial College London39
TL;DR: A comprehensive review of metal-halide perovskite nanocrystals can be found in this article, where researchers having expertise in different fields (chemistry, physics, and device engineering) have joined together to provide a state-of-the-art overview and future prospects of metalhalide nanocrystal research.
Abstract: Metal-halide perovskites have rapidly emerged as one of the most promising materials of the 21st century, with many exciting properties and great potential for a broad range of applications, from photovoltaics to optoelectronics and photocatalysis. The ease with which metal-halide perovskites can be synthesized in the form of brightly luminescent colloidal nanocrystals, as well as their tunable and intriguing optical and electronic properties, has attracted researchers from different disciplines of science and technology. In the last few years, there has been a significant progress in the shape-controlled synthesis of perovskite nanocrystals and understanding of their properties and applications. In this comprehensive review, researchers having expertise in different fields (chemistry, physics, and device engineering) of metal-halide perovskite nanocrystals have joined together to provide a state of the art overview and future prospects of metal-halide perovskite nanocrystal research.
471 citations
TL;DR: The most up-to-date progress on TMN-based nanomaterials is comprehensively reviewed, focusing on geometric-st structure design, electronic-structure engineering, and applications in electrochemical energy conversion and storage, including electrocatalysis, supercapacitors, and rechargeable batteries.
Abstract: Transition metal nitrides (TMNs), by virtue of their unique electronic structure, high electrical conductivity, superior chemical stability, and excellent mechanical robustness, have triggered tremendous research interest over the past decade, and showed great potential for electrochemical energy conversion and storage. However, bulk TMNs usually suffer from limited numbers of active sites and sluggish ionic kinetics, and eventually ordinary electrochemical performance. Designing nanostructured TMNs with tailored morphology and good dispersity has proved an effective strategy to address these issues, which provides a larger specific surface area, more abundant active sites, and shorter ion and mass transport distances over the bulk counterparts. Herein, the most up-to-date progress on TMN-based nanomaterials is comprehensively reviewed, focusing on geometric-structure design, electronic-structure engineering, and applications in electrochemical energy conversion and storage, including electrocatalysis, supercapacitors, and rechargeable batteries. Finally, we outline the future challenges of TMN-based nanomaterials and their possible research directions beyond electrochemical energy applications.
461 citations
TL;DR: This paper provided a comprehensive review of more than 150 deep learning-based models for text classification developed in recent years, and discussed their technical contributions, similarities, and strengths, and provided a quantitative analysis of the performance of different deep learning models on popular benchmarks.
Abstract: Deep learning--based models have surpassed classical machine learning--based approaches in various text classification tasks, including sentiment analysis, news categorization, question answering, and natural language inference. In this article, we provide a comprehensive review of more than 150 deep learning--based models for text classification developed in recent years, and we discuss their technical contributions, similarities, and strengths. We also provide a summary of more than 40 popular datasets widely used for text classification. Finally, we provide a quantitative analysis of the performance of different deep learning models on popular benchmarks, and we discuss future research directions.
457 citations
TL;DR: An Attention-based Bidirectional CNN-RNN Deep Model (ABCDM) is proposed that achieves state-of-the-art results on both long review and short tweet polarity classification and is evaluated on sentiment polarity detection.
385 citations
TL;DR: This review summarizes the recent advances of NIR-II photothermal combinational theranostics pertinent to chemotherapy, immunotherapy, radiotherapy, and photodynamic, sonodynamic, chemodynamic, gene, gas, ionic, vascular and magnetothermal therapy.
Abstract: Second near-infrared photothermal therapy (NIR-II PTT, 1000-1500 nm) has recently emerged as a new phototherapeutic modality with the advantages of deeper penetration, less energy dissipation and minimal normal-tissue toxicity over traditional first NIR PTT (750-1000 nm). However, suboptimal photothermal conversion and limited therapeutic efficacy remain the major challenges for NIR-II PTT. With the convergence in materials science, nanomedicine and biology, multifunctional NIR-II photothermal inorganic or organic materials have been extensively developed to combine NIR-II PTT with other therapeutic modalities for improved efficacies in treating life-threatening diseases including cancer and infection. This review summarizes the recent advances of NIR-II photothermal combinational theranostics pertinent to chemotherapy, immunotherapy, radiotherapy, and photodynamic, sonodynamic, chemodynamic, gene, gas, ionic, vascular and magnetothermal therapy. Potential obstacles and perspectives for future research and clinical translation of this new theranostic modality are also discussed.
378 citations
01 May 2021
TL;DR: A robust antimony single-atom photocatalyst (Sb-SAPC, single Sb atoms dispersed on carbon nitride) for the synthesis of H2O2 in a simple water and oxygen mixture under visible light irradiation was achieved in this article.
Abstract: Artificial photosynthesis offers a promising strategy to produce hydrogen peroxide (H2O2)—an environmentally friendly oxidant and a clean fuel. However, the low activity and selectivity of the two-electron oxygen reduction reaction (ORR) in the photocatalytic process greatly restricts the H2O2 production efficiency. Here we show a robust antimony single-atom photocatalyst (Sb-SAPC, single Sb atoms dispersed on carbon nitride) for the synthesis of H2O2 in a simple water and oxygen mixture under visible light irradiation. An apparent quantum yield of 17.6% at 420 nm together with a solar-to-chemical conversion efficiency of 0.61% for H2O2 synthesis was achieved. On the basis of time-dependent density function theory calculations, isotopic experiments and advanced spectroscopic characterizations, the photocatalytic performance is ascribed to the notably promoted two-electron ORR by forming μ-peroxide at the Sb sites and highly concentrated holes at the neighbouring N atoms. The in situ generated O2 via water oxidation is rapidly consumed by ORR, leading to boosted overall reaction kinetics. Hydrogen peroxide is an interesting target for artificial photosynthesis, although its actual production via the two-electron oxygen reduction reaction remains limited. Now, a carbon nitride-supported antimony single atom photocatalyst has been developed with a superior performance for this process.
308 citations
TL;DR: In this paper, the authors explore the emerging opportunities brought by 6G technologies in IoT networks and applications, by conducting a holistic survey on the convergence of 6G and IoT, and highlight interesting research challenges and point out potential directions to spur further research in this promising area.
Abstract: The sixth generation (6G) wireless communication networks are envisioned to revolutionize customer services and applications via the Internet of Things (IoT) towards a future of fully intelligent and autonomous systems. In this article, we explore the emerging opportunities brought by 6G technologies in IoT networks and applications, by conducting a holistic survey on the convergence of 6G and IoT. We first shed light on some of the most fundamental 6G technologies that are expected to empower future IoT networks, including edge intelligence, reconfigurable intelligent surfaces, space-air-ground-underwater communications, Terahertz communications, massive ultra-reliable and low-latency communications, and blockchain. Particularly, compared to the other related survey papers, we provide an in-depth discussion of the roles of 6G in a wide range of prospective IoT applications via five key domains, namely Healthcare Internet of Things, Vehicular Internet of Things and Autonomous Driving, Unmanned Aerial Vehicles, Satellite Internet of Things, and Industrial Internet of Things. Finally, we highlight interesting research challenges and point out potential directions to spur further research in this promising area.
TL;DR: A systematic review under both scientometric and qualitative analysis is presented to present the current state of AI adoption in the context of CEM and discuss its future research trends.
23 Mar 2021
TL;DR: In this paper, the authors investigated the peak levels and dynamics of neutralising antibody waning and IgG avidity maturation over time, and correlate this with clinical parameters, cytokines, and T-cell responses.
Abstract: Summary Background Studies have found different waning rates of neutralising antibodies compared with binding antibodies against SARS-CoV-2. The impact of neutralising antibody waning rate at the individual patient level on the longevity of immunity remains unknown. We aimed to investigate the peak levels and dynamics of neutralising antibody waning and IgG avidity maturation over time, and correlate this with clinical parameters, cytokines, and T-cell responses. Methods We did a longitudinal study of patients who had recovered from COVID-19 up to day 180 post-symptom onset by monitoring changes in neutralising antibody levels using a previously validated surrogate virus neutralisation test. Changes in antibody avidities and other immune markers at different convalescent stages were determined and correlated with clinical features. Using a machine learning algorithm, temporal change in neutralising antibody levels was classified into five groups and used to predict the longevity of neutralising antibody-mediated immunity. Findings We approached 517 patients for participation in the study, of whom 288 consented for outpatient follow-up and collection of serial blood samples. 164 patients were followed up and had adequate blood samples collected for analysis, with a total of 546 serum samples collected, including 128 blood samples taken up to 180 days post-symptom onset. We identified five distinctive patterns of neutralising antibody dynamics as follows: negative, individuals who did not, at our intervals of sampling, develop neutralising antibodies at the 30% inhibition level (19 [12%] of 164 patients); rapid waning, individuals who had varying levels of neutralising antibodies from around 20 days after symptom onset, but seroreverted in less than 180 days (44 [27%] of 164 patients); slow waning, individuals who remained neutralising antibody-positive at 180 days post-symptom onset (46 [28%] of 164 patients); persistent, although with varying peak neutralising antibody levels, these individuals had minimal neutralising antibody decay (52 [32%] of 164 patients); and delayed response, a small group that showed an unexpected increase of neutralising antibodies during late convalescence (at 90 or 180 days after symptom onset; three [2%] of 164 patients). Persistence of neutralising antibodies was associated with disease severity and sustained level of pro-inflammatory cytokines, chemokines, and growth factors. By contrast, T-cell responses were similar among the different neutralising antibody dynamics groups. On the basis of the different decay dynamics, we established a prediction algorithm that revealed a wide range of neutralising antibody longevity, varying from around 40 days to many decades. Interpretation Neutralising antibody response dynamics in patients who have recovered from COVID-19 vary greatly, and prediction of immune longevity can only be accurately determined at the individual level. Our findings emphasise the importance of public health and social measures in the ongoing pandemic outbreak response, and might have implications for longevity of immunity after vaccination. Funding National Medical Research Council, Biomedical Research Council, and A*STAR, Singapore.
TL;DR: In this paper, metal-organic frameworks (MOFs) have been identified as versatile precursors or sacrificial templates for preparing functional materials as advanced electrodes or high-efficiency catalysts for electrochemical energy storage and conversion (EESC).
Abstract: With many apparent advantages including high surface area, tunable pore sizes and topologies, and diverse periodic organic-inorganic ingredients, metal-organic frameworks (MOFs) have been identified as versatile precursors or sacrificial templates for preparing functional materials as advanced electrodes or high-efficiency catalysts for electrochemical energy storage and conversion (EESC) In this Mini Review, we first briefly summarize the material design strategies to show the rich possibilities of the chemical compositions and physical structures of MOFs derivatives We next highlight the latest advances focusing on the composition/structure/performance relationship and discuss their practical applications in various EESC systems, such as supercapacitors, rechargeable batteries, fuel cells, water electrolyzers, and carbon dioxide/nitrogen reduction reactions Finally, we provide some of our own insights into the major challenges and prospective solutions of MOF-derived functional materials for EESC, hoping to shed some light on the future development of this highly exciting field
TL;DR: In this paper, the authors used high-resolution scanning tunnelling microscopy to discover an unconventional chiral charge order in a kagome material, KV3Sb5, with both a topological band structure and a superconducting ground state.
Abstract: Intertwining quantum order and non-trivial topology is at the frontier of condensed matter physics1–4. A charge-density-wave-like order with orbital currents has been proposed for achieving the quantum anomalous Hall effect5,6 in topological materials and for the hidden phase in cuprate high-temperature superconductors7,8. However, the experimental realization of such an order is challenging. Here we use high-resolution scanning tunnelling microscopy to discover an unconventional chiral charge order in a kagome material, KV3Sb5, with both a topological band structure and a superconducting ground state. Through both topography and spectroscopic imaging, we observe a robust 2 × 2 superlattice. Spectroscopically, an energy gap opens at the Fermi level, across which the 2 × 2 charge modulation exhibits an intensity reversal in real space, signalling charge ordering. At the impurity-pinning-free region, the strength of intrinsic charge modulations further exhibits chiral anisotropy with unusual magnetic field response. Theoretical analysis of our experiments suggests a tantalizing unconventional chiral charge density wave in the frustrated kagome lattice, which can not only lead to a large anomalous Hall effect with orbital magnetism, but also be a precursor of unconventional superconductivity. An unconventional chiral charge order is observed in a kagome superconductor by scanning tunnelling microscopy. This charge order has unusual magnetic tunability and intertwines with electronic topology.
TL;DR: A federated learning (FL) system leveraging a reputation mechanism to assist home appliance manufacturers to train a machine learning model based on customers’ data so that manufacturers can predict customers' requirements and consumption behaviors in the future.
Abstract: Home appliance manufacturers strive to obtain feedback from users to improve their products and services to build a smart home system. To help manufacturers develop a smart home system, we design a federated learning (FL) system leveraging a reputation mechanism to assist home appliance manufacturers to train a machine learning model based on customers’ data. Then, manufacturers can predict customers’ requirements and consumption behaviors in the future. The working flow of the system includes two stages: in the first stage, customers train the initial model provided by the manufacturer using both the mobile phone and the mobile-edge computing (MEC) server. Customers collect data from various home appliances using phones, and then they download and train the initial model with their local data. After deriving local models, customers sign on their models and send them to the blockchain. In case customers or manufacturers are malicious, we use the blockchain to replace the centralized aggregator in the traditional FL system. Since records on the blockchain are untampered, malicious customers or manufacturers’ activities are traceable. In the second stage, manufacturers select customers or organizations as miners for calculating the averaged model using received models from customers. By the end of the crowdsourcing task, one of the miners, who is selected as the temporary leader, uploads the model to the blockchain. To protect customers’ privacy and improve the test accuracy, we enforce differential privacy (DP) on the extracted features and propose a new normalization technique. We experimentally demonstrate that our normalization technique outperforms batch normalization when features are under DP protection. In addition, to attract more customers to participate in the crowdsourcing FL task, we design an incentive mechanism to award participants.
TL;DR: The impact of SARS-CoV-2 variants of concern (VOCs) on disease severity is unclear as discussed by the authors, and the impact of vaccination status was associated with decreased severity.
Abstract: Background he impact of SARS-CoV-2 variants of concern (VOCs) on disease severity is unclear. In this retrospective study, we compared outcomes of patients infected with B.1.1.7, B.1.351, and B.1.617.2 with those with wild-type strains from early 2020. Methods National surveillance data from 1-January-2021 to 22-May-2021 were obtained from the Ministry of Health, and outcomes in relation to VOC were explored. Detailed patient level data from all patients with VOC infection admitted to our center between 20-December-2020 and 12-May-2021 were analyzed. Clinical outcomes were compared with a cohort of 846 patients admitted from January-April 2020. Results 829 patients in Singapore in the study period were infected with these 3 VOCs. After adjusting for age and sex, B.1.617.2 was associated with higher odds of oxygen requirement, ICU admission, or death (adjusted odds ratio (aOR) 4.90, [95% CI 1.43-30.78]). 157 of these patients were admitted to our center. After adjusting for age, sex, comorbidities, and vaccination, aOR for pneumonia with B.1.617.2 was 1.88 [95% CI 0.95-3.76]) compared with wild-type. These differences were not seen with B.1.1.7 and B.1.351. Vaccination status was associated with decreased severity. B.1.617.2 was associated with significantly lower PCR Ct values and longer duration of Ct value ≤30 (median duration 18 days for B.1.617.2, 13 days for wild-type). Conclusions There was a signal toward increased severity associated with B.1.617.2. The association of B.1.617.2 with lower Ct value and longer viral shedding provides a potential mechanism for increased transmissibility. These findings provide an impetus for the rapid implementation of vaccination programs.
Max Planck Society1, Weizmann Institute of Science2, Ben-Gurion University of the Negev3, Stanford University4, Leibniz Association5, University of Jena6, University of Tübingen7, Royal Holloway, University of London8, Wageningen University and Research Centre9, University of Groningen10, Hainan University11, Chalmers University of Technology12, Chiba University13, ETH Zurich14, Cornell University15, Scripps Research Institute16, University of Missouri17, Fudan University18, Nara Institute of Science and Technology19, Nanyang Technological University20, Huazhong Agricultural University21, Dalian Institute of Chemical Physics22
TL;DR: In this article, the authors present guidelines covering sample preparation, replication and randomization, quantification, recovery and recombination, ion suppression and peak misidentification, as a means to enable high-quality reporting of liquid chromatography and gas chromatography-mass spectrometry-derived data.
Abstract: Mass spectrometry-based metabolomics approaches can enable detection and quantification of many thousands of metabolite features simultaneously. However, compound identification and reliable quantification are greatly complicated owing to the chemical complexity and dynamic range of the metabolome. Simultaneous quantification of many metabolites within complex mixtures can additionally be complicated by ion suppression, fragmentation and the presence of isomers. Here we present guidelines covering sample preparation, replication and randomization, quantification, recovery and recombination, ion suppression and peak misidentification, as a means to enable high-quality reporting of liquid chromatography- and gas chromatography-mass spectrometry-based metabolomics-derived data.
TL;DR: In this paper, the authors provide the current progress on understanding of OER mechanisms in acid, analyze the promising strategies to enhance both activity and stability, and summarize the state-of-the-art catalysts for OER in acid.
Abstract: The proton exchange membrane (PEM) water electrolysis is one of the most promising hydrogen production techniques. The oxygen evolution reaction (OER) occurring at the anode dominates the overall efficiency. Developing active and robust electrocatalysts for OER in acid is a longstanding challenge for PEM water electrolyzers. Most catalysts show unsatisfied stability under strong acidic and oxidative conditions. Such a stability challenge also leads to difficulties for a better understanding of mechanisms. This review aims to provide the current progress on understanding of OER mechanisms in acid, analyze the promising strategies to enhance both activity and stability, and summarize the state-of-the-art catalysts for OER in acid. First, the prevailing OER mechanisms are reviewed to establish the physicochemical structure-activity relationships for guiding the design of highly efficient OER electrocatalysts in acid with stable performance. The reported approaches to improve the activity, from macroview to microview, are then discussed. To analyze the problem of instability, the key factors affecting catalyst stability are summarized and the surface reconstruction is discussed. Various noble-metal-based OER catalysts and the current progress of non-noble-metal-based catalysts are reviewed. Finally, the challenges and perspectives for the development of active and robust OER catalysts in acid are discussed.
TL;DR: Several main issues in FLchain design are identified, including communication cost, resource allocation, incentive mechanism, security and privacy protection, and the applications of FLchain in popular MEC domains, such as edge data sharing, edge content caching and edge crowdsensing are investigated.
Abstract: Mobile-edge computing (MEC) has been envisioned as a promising paradigm to handle the massive volume of data generated from ubiquitous mobile devices for enabling intelligent services with the help of artificial intelligence (AI). Traditionally, AI techniques often require centralized data collection and training in a single entity, e.g., an MEC server, which is now becoming a weak point due to data privacy concerns and high overhead of raw data communications. In this context, federated learning (FL) has been proposed to provide collaborative data training solutions, by coordinating multiple mobile devices to train a shared AI model without directly exposing their underlying data, which enjoys considerable privacy enhancement. To improve the security and scalability of FL implementation, blockchain as a ledger technology is attractive for realizing decentralized FL training without the need for any central server. Particularly, the integration of FL and blockchain leads to a new paradigm, called FLchain , which potentially transforms intelligent MEC networks into decentralized, secure, and privacy-enhancing systems. This article presents an overview of the fundamental concepts and explores the opportunities of FLchain in MEC networks. We identify several main issues in FLchain design, including communication cost, resource allocation, incentive mechanism, security and privacy protection. The key solutions and the lessons learned along with the outlooks are also discussed. Then, we investigate the applications of FLchain in popular MEC domains, such as edge data sharing, edge content caching and edge crowdsensing. Finally, important research challenges and future directions are also highlighted.
TL;DR: In the authors’ perspective, in situ monitoring of AM processes will significantly benefit from the object detection ability of ML, and data sharing of AM would enable faster adoption of ML in AM.
Abstract: Additive manufacturing (AM) or 3D printing is growing rapidly in the manufacturing industry and has gained a lot of attention from various fields owing to its ability to fabricate parts with complex features. The reliability of the 3D printed parts has been the focus of the researchers to realize AM as an end-part production tool. Machine learning (ML) has been applied in various aspects of AM to improve the whole design and manufacturing workflow especially in the era of industry 4.0. In this review article, various types of ML techniques are first introduced. It is then followed by the discussion on their use in various aspects of AM such as design for 3D printing, material tuning, process optimization, in situ monitoring, cloud service, and cybersecurity. Potential applications in the biomedical, tissue engineering and building and construction will be highlighted. The challenges faced by ML in AM such as computational cost, standards for qualification and data acquisition techniques will also be discussed. In the authors’ perspective, in situ monitoring of AM processes will significantly benefit from the object detection ability of ML. As a large data set is crucial for ML, data sharing of AM would enable faster adoption of ML in AM. Standards for the shared data are needed to facilitate easy sharing of data. The use of ML in AM will become more mature and widely adopted as better data acquisition techniques and more powerful computer chips for ML are developed.
TL;DR: In this article, the authors focus on hydrogen electrification through proton exchange membrane fuel cells (PEMFCs), which are widely believed to be commercially suitable for automotive applications, particularly for vehicles requiring minimal hydrogen infrastructure support, such as fleets of taxies, buses, and logistic vehicles.
TL;DR: NeuroKit2 as discussed by the authors is an open-source, community-driven, and user-centered Python package for neurophysiological signal processing, which includes high-level functions that enable data processing in a few lines of code using validated pipelines.
Abstract: NeuroKit2 is an open-source, community-driven, and user-centered Python package for neurophysiological signal processing. It provides a comprehensive suite of processing routines for a variety of bodily signals (e.g., ECG, PPG, EDA, EMG, RSP). These processing routines include high-level functions that enable data processing in a few lines of code using validated pipelines, which we illustrate in two examples covering the most typical scenarios, such as an event-related paradigm and an interval-related analysis. The package also includes tools for specific processing steps such as rate extraction and filtering methods, offering a trade-off between high-level convenience and fine-tuned control. Its goal is to improve transparency and reproducibility in neurophysiological research, as well as foster exploration and innovation. Its design philosophy is centred on user-experience and accessibility to both novice and advanced users.
TL;DR: In this paper, a photothermally activatable polymeric pro-nanoagonist (APNA) was constructed from covalent conjugation of an immunostimulant onto a NIR-II semiconducting transducer through a labile thermo-responsive linker, which not only triggers tumor ablation and immunogenic cell death but also initiates the cleavage of thermolabile linker to liberate caged agonist for in-situ immune activation in deep solid tumor (8 mm).
Abstract: Nanomedicine in combination with immunotherapy offers opportunities to treat cancer in a safe and effective manner; however, remote control of immune response with spatiotemporal precision remains challenging We herein report a photothermally activatable polymeric pro-nanoagonist (APNA) that is specifically regulated by deep-tissue-penetrating second near-infrared (NIR-II) light for combinational photothermal immunotherapy APNA is constructed from covalent conjugation of an immunostimulant onto a NIR-II semiconducting transducer through a labile thermo-responsive linker Upon NIR-II photoirradiation, APNA mediates photothermal effect, which not only triggers tumor ablation and immunogenic cell death but also initiates the cleavage of thermolabile linker to liberate caged agonist for in-situ immune activation in deep solid tumor (8 mm) Such controlled immune regulation potentiates systemic antitumor immunity, leading to promoted cytotoxic T lymphocytes and helper T cell infiltration in distal tumor, lung and liver to inhibit cancer metastasis Thereby, the present work illustrates a generic strategy to prepare pro-immunostimulants for spatiotemporal regulation of cancer nano-immunotherapy Precise control of immune response remains challenging for cancer immunotherapy Here, the authors report on photothermally activatable semiconducting polymeric pro-agonist in response to second near-infrared window light for regulated photothermal immunotherapy
TL;DR: In this paper, the fundamental natures of heterostructures, including charge redistribution, built-in electric field, and associated energy storage mechanisms are summarized and discussed in detail, and various synthesis routes for heterostructure in energy storage fields are roundly reviewed, and their advantages and drawbacks are analyzed.
Abstract: With the ever-increasing adaption of large-scale energy storage systems and electric devices, the energy storage capability of batteries and supercapacitors has faced increased demand and challenges. The electrodes of these devices have experienced radical change with the introduction of nano-scale materials. As new generation materials, heterostructure materials have attracted increasing attention due to their unique interfaces, robust architectures, and synergistic effects, and thus, the ability to enhance the energy/power outputs as well as the lifespan of batteries. In this review, the recent progress in heterostructure from energy storage fields is summarized. Specifically, the fundamental natures of heterostructures, including charge redistribution, built-in electric field, and associated energy storage mechanisms, are summarized and discussed in detail. Furthermore, various synthesis routes for heterostructures in energy storage fields are roundly reviewed, and their advantages and drawbacks are analyzed. The superiorities and current achievements of heterostructure materials in lithium-ion batteries (LIBs), sodium-ion batteries (SIBs), lithium-sulfur batteries (Li-S batteries), supercapacitors, and other energy storage devices are discussed. Finally, the authors conclude with the current challenges and perspectives of the heterostructure materials for the fields of energy storage.
TL;DR: Effective integration between the electronic components with garments, human skin, and living organisms is illustrated, presenting multifunctional platforms with self-powered potential for human-robot interactions and biomedicine.
Abstract: Soft robotics inspired by the movement of living organisms, with excellent adaptability and accuracy for accomplishing tasks, are highly desirable for efficient operations and safe interactions with human. With the emerging wearable electronics, higher tactility and skin affinity are pursued for safe and user-friendly human-robot interactions. Fabrics interlocked by fibers perform traditional static functions such as warming, protection, and fashion. Recently, dynamic fibers and fabrics are favorable to deliver active stimulus responses such as sensing and actuating abilities for soft-robots and wearables. First, the responsive mechanisms of fiber/fabric actuators and their performances under various external stimuli are reviewed. Fiber/yarn-based artificial muscles for soft-robots manipulation and assistance in human motion are discussed, as well as smart clothes for improving human perception. Second, the geometric designs, fabrications, mechanisms, and functions of fibers/fabrics for sensing and energy harvesting from the human body and environments are summarized. Effective integration between the electronic components with garments, human skin, and living organisms is illustrated, presenting multifunctional platforms with self-powered potential for human-robot interactions and biomedicine. Lastly, the relationships between robotic/wearable fibers/fabrics and the external stimuli, together with the challenges and possible routes for revolutionizing the robotic fibers/fabrics and wearables in this new era are proposed.
28 Apr 2021
TL;DR: In this paper, an attention-based deep learning architecture called AttnSleep was proposed to classify sleep stages using single-channel EEG signals, which leverages a multi-head attention mechanism to capture the temporal dependencies among the extracted features.
Abstract: Automatic sleep stage mymargin classification is of great importance to measure sleep quality. In this paper, we propose a novel attention-based deep learning architecture called AttnSleep to classify sleep stages using single channel EEG signals. This architecture starts with the feature extraction module based on multi-resolution convolutional neural network (MRCNN) and adaptive feature recalibration (AFR). The MRCNN can extract low and high frequency features and the AFR is able to improve the quality of the extracted features by modeling the inter-dependencies between the features. The second module is the temporal context encoder (TCE) that leverages a multi-head attention mechanism to capture the temporal dependencies among the extracted features. Particularly, the multi-head attention deploys causal convolutions to model the temporal relations in the input features. We evaluate the performance of our proposed AttnSleep model using three public datasets. The results show that our AttnSleep outperforms state-of-the-art techniques in terms of different evaluation metrics. Our source codes, experimental data, and supplementary materials are available at https://github.com/emadeldeen24/AttnSleep .
TL;DR: In this article, a DRL-based secure beamforming approach was proposed to achieve the optimal beamforming policy against eavesdroppers in dynamic environments, and a modified postdecision state (PDS) and prioritized experience replay (PER) scheme was utilized to enhance the learning efficiency and secrecy performance.
Abstract: In this paper, we study an intelligent reflecting surface (IRS)-aided wireless secure communication system, where an IRS is deployed to adjust its reflecting elements to secure the communication of multiple legitimate users in the presence of multiple eavesdroppers. Aiming to improve the system secrecy rate, a design problem for jointly optimizing the base station (BS)’s beamforming and the IRS’s reflecting beamforming is formulated considering different quality of service (QoS) requirements and time-varying channel conditions. As the system is highly dynamic and complex, and it is challenging to address the non-convex optimization problem, a novel deep reinforcement learning (DRL)-based secure beamforming approach is firstly proposed to achieve the optimal beamforming policy against eavesdroppers in dynamic environments. Furthermore, post-decision state (PDS) and prioritized experience replay (PER) schemes are utilized to enhance the learning efficiency and secrecy performance. Specifically, a modified PDS scheme is presented to trace the channel dynamic and adjust the beamforming policy against channel uncertainty accordingly. Simulation results demonstrate that the proposed deep PDS-PER learning based secure beamforming approach can significantly improve the system secrecy rate and QoS satisfaction probability in IRS-aided secure communication systems.
TL;DR: There was a signal toward increased severity associated with B.1.617.2 with lower Ct value and longer viral shedding providing a potential mechanism for increased transmissibility, and these findings provide a strong impetus for the rapid implementation of vaccination programmes.
Abstract: Background: The impact of SARS-CoV-2 variants of concern (VOCs) on disease severity is unclear. In this retrospective cohort study, we compared outcomes of patients infected with B.1.1.7, B.1.351, and B.1.617.2 with those with wild-type strains from early 2020.
Methods: National surveillance data from 1-January-2021 to 22-May-2021 were obtained from the Ministry of Health, and outcomes in relation to VOC were explored. Detailed patient level data from all SARS-CoV-2 patients with VOC infection admitted to our centre between 20-December-2020 and 12-May-2021 were analysed. Outcomes were compared with a cohort of 846 patients admitted January-April 2020.
Findings: There were 838 VOC infections in Singapore in the study period. After adjusting for age and gender, B.1.617.2 infection was associated with higher odds of oxygen requirement, ICU admission, or death (adjusted odds ratio (aOR) 4·90, [95% CI 1·43-30·78]. 157 patients with VOCs were admitted to our centre. After adjusting for age, gender, comorbidities, and vaccination, aOR for pneumonia with B.1.617.2 was 1·88 [95% CI 0·95-3·76]) compared with wild-type. B.1.617.2 was associated with significantly lower PCR Ct values and significantly longer duration of Ct value ≤30 (estimated median duration 18 days for B.1.617.2, 13 days for wild-type). Vaccine breakthrough cases were less severe.
Interpretation: There was a signal toward increased severity associated with B.1.617.2. The association of B.1.617.2 with lower Ct value and longer viral shedding provides a potential mechanism for increased transmissibility. These findings provide a strong impetus for the rapid implementation of vaccination programmes.
Funding Information: National Medical Research Council grants COVID19RF-001 and COVID19RF-008.
Declaration of Interests: BEY reports personal fees from Roche and Sanofi, outside the submitted work. All other authors declare no competing interests.
Ethics Approval Statement: Informed consent for retrospective data collection was waived as approved by the institutional review board (NHG-DSRB reference number 2020/01122).
TL;DR: This review highlights how a specific chemical, physical and optical property of 2D materials can influence the performance of bio/sensing, improve drug delivery and photo/thermal therapy as well as affect their toxicity.
Abstract: Two-dimensional (2D) materials are at the forefront of materials research. Here we overview their applications beyond graphene, such as transition metal dichalcogenides, monoelemental Xenes (including phosphorene and bismuthene), carbon nitrides, boron nitrides along with transition metal carbides and nitrides (MXenes). We discuss their usage in various biomedical and environmental monitoring applications, from biosensors to therapeutic treatment agents, their toxicity and their utility in chemical sensing. We highlight how a specific chemical, physical and optical property of 2D materials can influence the performance of bio/sensing, improve drug delivery and photo/thermal therapy as well as affect their toxicity. Such properties are determined by crystal phases electrical conductivity, degree of exfoliation, surface functionalization, strong photoluminescence, strong optical absorption in the near-infrared range and high photothermal conversion efficiency. This review conveys the great future of all the families of 2D materials, especially with the expanding 2D materials' landscape as new materials emerge such as germanene and silicene.