Showing papers by "Rensselaer Polytechnic Institute published in 2019"
532 citations
Columbia University1, Stockholm University2, University of Bologna3, University of Mainz4, University of Münster5, University of Coimbra6, University of Turin7, New York University Abu Dhabi8, University of Zurich9, Rensselaer Polytechnic Institute10, University of Amsterdam11, Max Planck Society12, Weizmann Institute of Science13, University of Freiburg14, University of Nantes15, Purdue University16, University of California, San Diego17, University of Chicago18, Nagoya University19, Pierre-and-Marie-Curie University20, Université Paris-Saclay21, Rice University22, University of California, Los Angeles23
TL;DR: Constraints on light dark matter (DM) models using ionization signals in the XENON1T experiment are reported, and no DM or CEvNS detection may be claimed because the authors cannot model all of their backgrounds.
Abstract: We report constraints on light dark matter (DM) models using ionization signals in the XENON1T experiment. We mitigate backgrounds with strong event selections, rather than requiring a scintillation signal, leaving an effective exposure of (22±3) tonne day. Above ∼0.4 keVee, we observe 30 MeV/c2, and absorption of dark photons and axionlike particles for mχ within 0.186–1 keV/c2.
412 citations
University of California, Riverside1, Purdue University2, Pacific Northwest National Laboratory3, Rensselaer Polytechnic Institute4, SUNY Downstate Medical Center5, University of Michigan6, Brown University7, University of Pennsylvania8, University of California, Santa Barbara9, Stanford University10
TL;DR: It is demonstrated that machine learning and multiscale modeling can naturally complement each other to create robust predictive models that integrate the underlying physics to manage ill-posed problems and explore massive design spaces.
Abstract: Fueled by breakthrough technology developments, the biological, biomedical, and behavioral sciences are now collecting more data than ever before. There is a critical need for time- and cost-efficient strategies to analyze and interpret these data to advance human health. The recent rise of machine learning as a powerful technique to integrate multimodality, multifidelity data, and reveal correlations between intertwined phenomena presents a special opportunity in this regard. However, machine learning alone ignores the fundamental laws of physics and can result in ill-posed problems or non-physical solutions. Multiscale modeling is a successful strategy to integrate multiscale, multiphysics data and uncover mechanisms that explain the emergence of function. However, multiscale modeling alone often fails to efficiently combine large datasets from different sources and different levels of resolution. Here we demonstrate that machine learning and multiscale modeling can naturally complement each other to create robust predictive models that integrate the underlying physics to manage ill-posed problems and explore massive design spaces. We review the current literature, highlight applications and opportunities, address open questions, and discuss potential challenges and limitations in four overarching topical areas: ordinary differential equations, partial differential equations, data-driven approaches, and theory-driven approaches. Towards these goals, we leverage expertise in applied mathematics, computer science, computational biology, biophysics, biomechanics, engineering mechanics, experimentation, and medicine. Our multidisciplinary perspective suggests that integrating machine learning and multiscale modeling can provide new insights into disease mechanisms, help identify new targets and treatment strategies, and inform decision making for the benefit of human health.
315 citations
TL;DR: In this article, a large-scale assessment of lake ice loss is presented, using observations from 513 lakes around the Northern Hemisphere, revealing the importance of air temperature, lake depth, elevation and shoreline complexity in governing ice cover.
Abstract: Ice provides a range of ecosystem services—including fish harvest1, cultural traditions2, transportation3, recreation4 and regulation of the hydrological cycle5—to more than half of the world’s 117 million lakes. One of the earliest observed impacts of climatic warming has been the loss of freshwater ice6, with corresponding climatic and ecological consequences7. However, while trends in ice cover phenology have been widely documented2,6,8,9, a comprehensive large-scale assessment of lake ice loss is absent. Here, using observations from 513 lakes around the Northern Hemisphere, we identify lakes vulnerable to ice-free winters. Our analyses reveal the importance of air temperature, lake depth, elevation and shoreline complexity in governing ice cover. We estimate that 14,800 lakes currently experience intermittent winter ice cover, increasing to 35,300 and 230,400 at 2 and 8 °C, respectively, and impacting up to 394 and 656 million people. Our study illustrates that an extensive loss of lake ice will occur within the next generation, stressing the importance of climate mitigation strategies to preserve ecosystem structure and function, as well as local winter cultural heritage. Up to 35,000 lakes in the Northern Hemisphere may be at risk of intermittent winter ice cover at 2 °C warming, reveals an observation-based study. This would affect 394 million people reliant on lake ice for ecosystem services.
274 citations
TL;DR: In this article, a comprehensive assessment of recent studies on the removal of various contaminants of emerging concern (CECs) with different physicochemical properties by various MOF-NAs under various water quality conditions (e.g., pH, background ions/ionic strength, natural organic matter, and temperature).
270 citations
TL;DR: In this article, a modularized neural network for low-dose CT (LDCT) was proposed and compared with commercial iterative reconstruction methods from three leading CT vendors, and the learned workflow allows radiologists-in-the-loop to optimize the denoising depth in a task-specific fashion.
Abstract: Commercial iterative reconstruction techniques help to reduce CT radiation dose but altered image appearance and artifacts limit their adoptability and potential use. Deep learning has been investigated for low-dose CT (LDCT). Here we design a modularized neural network for LDCT and compared it with commercial iterative reconstruction methods from three leading CT vendors. While popular networks are trained for an end-to-end mapping, our network performs an end-to-process mapping so that intermediate denoised images are obtained with associated noise reduction directions towards a final denoised image. The learned workflow allows radiologists-in-the-loop to optimize the denoising depth in a task-specific fashion. Our network was trained with the Mayo LDCT Dataset, and tested on separate chest and abdominal CT exams from Massachusetts General Hospital. The best deep learning reconstructions were systematically compared to the best iterative reconstructions in a double-blinded reader study. This study confirms that our deep learning approach performed either favorably or comparably in terms of noise suppression and structural fidelity, and is much faster than the commercial iterative reconstruction algorithms.
265 citations
Tsinghua University1, University of Central Florida2, Iowa State University3, Southern California Edison4, Rensselaer Polytechnic Institute5, Southern Methodist University6, Microsoft7, South Dakota State University8, Federal University of Rio de Janeiro9, ETH Zurich10, Technical University of Denmark11
TL;DR: The motivation, state-of-the-art, and future directions of the coordination of transmission system operators (TSO) and distribution system operator (DSO) are thoroughly discussed.
Abstract: In this paper, we review the emerging challenges and research opportunities for voltage control in smart grids. For transmission grids, the voltage control for accommodating wind and solar power, fault-induced delayed voltage recovery, and measurement-based Thevenin equivalent for voltage stability analysis are reviewed. For distribution grids, the impact of high penetration of distributed energy resources is analyzed, typical control strategies are reviewed, and the challenges for local inverter Volt–Var control is discussed. In addition, the motivation, state-of-the-art, and future directions of the coordination of transmission system operators (TSO) and distribution system operators (DSO) are also thoroughly discussed.
246 citations
Columbia University1, Stockholm University2, University of Bologna3, University of Mainz4, University of Münster5, University of Coimbra6, New York University Abu Dhabi7, University of Zurich8, Rensselaer Polytechnic Institute9, University of Amsterdam10, Max Planck Society11, Weizmann Institute of Science12, University of Freiburg13, University of Nantes14, University of California, San Diego15, University of Chicago16, Nagoya University17, Purdue University18, Pierre-and-Marie-Curie University19, Université Paris-Saclay20, Rice University21, University of California, Los Angeles22
TL;DR: The analysis uses the full ton year exposure of XENON1T to constrain the spin-dependent proton-only and neutron-only cases and sets exclusion limits on the WIMP-nucleon interactions.
Abstract: We report the first experimental results on spin-dependent elastic weakly interacting massive particle (WIMP) nucleon scattering from the XENON1T dark matter search experiment. The analysis uses the full ton year exposure of XENON1T to constrain the spin-dependent proton-only and neutron-only cases. No significant signal excess is observed, and a profile likelihood ratio analysis is used to set exclusion limits on the WIMP-nucleon interactions. This includes the most stringent constraint to date on the WIMP-neutron cross section, with a minimum of 6.3×10-42 cm2 at 30 GeV/c2 and 90% confidence level. The results are compared with those from collider searches and used to exclude new parameter space in an isoscalar theory with an axial-vector mediator.
241 citations
TL;DR: This research roadmap is intended to identify and prioritize needs for academic research laboratories, funding agencies, professional societies, and industry to facilitate wide availability of clinical imaging data sets.
Abstract: Imaging research laboratories are rapidly creating machine learning systems that achieve expert human performance using open-source methods and tools. These artificial intelligence systems are being developed to improve medical image reconstruction, noise reduction, quality assurance, triage, segmentation, computer-aided detection, computer-aided classification, and radiogenomics. In August 2018, a meeting was held in Bethesda, Maryland, at the National Institutes of Health to discuss the current state of the art and knowledge gaps and to develop a roadmap for future research initiatives. Key research priorities include: 1, new image reconstruction methods that efficiently produce images suitable for human interpretation from source data; 2, automated image labeling and annotation methods, including information extraction from the imaging report, electronic phenotyping, and prospective structured image reporting; 3, new machine learning methods for clinical imaging data, such as tailored, pretrained model architectures, and federated machine learning methods; 4, machine learning methods that can explain the advice they provide to human users (so-called explainable artificial intelligence); and 5, validated methods for image de-identification and data sharing to facilitate wide availability of clinical imaging data sets. This research roadmap is intended to identify and prioritize these needs for academic research laboratories, funding agencies, professional societies, and industry.
228 citations
15 Jun 2019
TL;DR: In this paper, a new attention-driven Siamese learning architecture, called Consistent Attentive siamese Network (CASN), is proposed for person re-ID.
Abstract: We propose a new deep architecture for person re-identification (re-id). While re-id has seen much recent progress, spatial localization and view-invariant representation learning for robust cross-view matching remain key, unsolved problems. We address these questions by means of a new attention-driven Siamese learning architecture, called the Consistent Attentive Siamese Network. Our key innovations compared to existing, competing methods include (a) a flexible framework design that produces attention with only identity labels as supervision, (b) explicit mechanisms to enforce attention consistency among images of the same person, and (c) a new Siamese framework that integrates attention and attention consistency, producing principled supervisory signals as well as the first mechanism that can explain the reasoning behind the Siamese framework’s predictions. We conduct extensive evaluations on the CUHK03-NP, DukeMTMC-ReID, and Market-1501 datasets and report competitive performance.
215 citations
TL;DR: This paper performs an extensive review of the facial landmark detection algorithms and identifies future research directions, including combining methods in different categories to leverage their respective strengths to solve landmark detection “in-the-wild”.
Abstract: The locations of the fiducial facial landmark points around facial components and facial contour capture the rigid and non-rigid facial deformations due to head movements and facial expressions. They are hence important for various facial analysis tasks. Many facial landmark detection algorithms have been developed to automatically detect those key points over the years, and in this paper, we perform an extensive review of them. We classify the facial landmark detection algorithms into three major categories: holistic methods, Constrained Local Model (CLM) methods, and the regression-based methods. They differ in the ways to utilize the facial appearance and shape information. The holistic methods explicitly build models to represent the global facial appearance and shape information. The CLMs explicitly leverage the global shape model but build the local appearance models. The regression based methods implicitly capture facial shape and appearance information. For algorithms within each category, we discuss their underlying theories as well as their differences. We also compare their performances on both controlled and in the wild benchmark datasets, under varying facial expressions, head poses, and occlusion. Based on the evaluations, we point out their respective strengths and weaknesses. There is also a separate section to review the latest deep learning based algorithms. The survey also includes a listing of the benchmark databases and existing software. Finally, we identify future research directions, including combining methods in different categories to leverage their respective strengths to solve landmark detection "in-the-wild".
TL;DR: The fabricated MXene@CS@PU sponge-based sensor has high compressibility and stable piezoresistive response for compressive strains of up to 85% with a stress of 245.7 kPa, and it also exhibits a satisfactory reproducibility for 5000 compression-release cycles.
Posted Content•
TL;DR: This paper introduces Cosmos QA, a large-scale dataset of 35,600 problems that require commonsense-based reading comprehension, formulated as multiple-choice questions, and proposes a new architecture that improves over the competitive baselines.
Abstract: Understanding narratives requires reading between the lines, which in turn, requires interpreting the likely causes and effects of events, even when they are not mentioned explicitly. In this paper, we introduce Cosmos QA, a large-scale dataset of 35,600 problems that require commonsense-based reading comprehension, formulated as multiple-choice questions. In stark contrast to most existing reading comprehension datasets where the questions focus on factual and literal understanding of the context paragraph, our dataset focuses on reading between the lines over a diverse collection of people's everyday narratives, asking such questions as "what might be the possible reason of ...?", or "what would have happened if ..." that require reasoning beyond the exact text spans in the context. To establish baseline performances on Cosmos QA, we experiment with several state-of-the-art neural architectures for reading comprehension, and also propose a new architecture that improves over the competitive baselines. Experimental results demonstrate a significant gap between machine (68.4%) and human performance (94%), pointing to avenues for future research on commonsense machine comprehension. Dataset, code and leaderboard is publicly available at this https URL.
31 Aug 2019
TL;DR: Cosmos QA as discussed by the authors ) is a large-scale dataset of 35,600 problems that require commonsense-based reading comprehension, formulated as multiple-choice questions, where the questions focus on reading between the lines, which in turn requires interpreting the likely causes and effects of events.
Abstract: Understanding narratives requires reading between the lines, which in turn, requires interpreting the likely causes and effects of events, even when they are not mentioned explicitly. In this paper, we introduce Cosmos QA, a large-scale dataset of 35,600 problems that require commonsense-based reading comprehension, formulated as multiple-choice questions. In stark contrast to most existing reading comprehension datasets where the questions focus on factual and literal understanding of the context paragraph, our dataset focuses on reading between the lines over a diverse collection of people’s everyday narratives, asking such questions as “what might be the possible reason of ...?", or “what would have happened if ..." that require reasoning beyond the exact text spans in the context. To establish baseline performances on Cosmos QA, we experiment with several state-of-the-art neural architectures for reading comprehension, and also propose a new architecture that improves over the competitive baselines. Experimental results demonstrate a significant gap between machine (68.4%) and human performance (94%), pointing to avenues for future research on commonsense machine comprehension. Dataset, code and leaderboard is publicly available at https://wilburone.github.io/cosmos.
TL;DR: Graphene structures with tunable graphene sheet alignment and orientation, obtained via microfluidic design, are reported, enabling strong size and geometry confinements and control over flow patterns.
Abstract: Macroscopic graphene structures such as graphene papers and fibres can be manufactured from individual two-dimensional graphene oxide sheets by a fluidics-enabled assembling process. However, achieving high thermal-mechanical and electrical properties is still challenging due to non-optimized microstructures and morphology. Here, we report graphene structures with tunable graphene sheet alignment and orientation, obtained via microfluidic design, enabling strong size and geometry confinements and control over flow patterns. Thin flat channels can be used to fabricate macroscopic graphene structures with perfectly stacked sheets that exhibit superior thermal and electrical conductivities and improved mechanical strength. We attribute the observed shape and size confinements to the flat distribution of shear stress from the anisotropic microchannel walls and the enhanced shear thinning degree of large graphene oxide sheets in solution. Elongational and step expansion flows are created to produce large-scale graphene tubes and rods with horizontally and perpendicularly aligned graphene sheets by tuning the elongational and extensional shear rates, respectively. Sheet alignment and orientation order of graphene structures induced by microfluidics design enable the optimization of electronic and mechanical properties of macroscopic graphene fibres.
Columbia University1, Stockholm University2, University of Bologna3, University of Mainz4, University of Münster5, University of Coimbra6, University of Turin7, New York University Abu Dhabi8, University of Zurich9, Rensselaer Polytechnic Institute10, University of Amsterdam11, Max Planck Society12, Weizmann Institute of Science13, University of Freiburg14, University of Nantes15, Purdue University16, University of California, San Diego17, University of Chicago18, Nagoya University19, Pierre-and-Marie-Curie University20, Université Paris-Saclay21, Rice University22, University of California, Los Angeles23
TL;DR: A probe of low-mass dark matter with masses down to about 85 MeV/c^{2} is reported on by looking for electronic recoils induced by the Migdal effect and bremsstrahlung using data from the XENON1T experiment, and exploiting an approach that uses ionization signals only allows for a lower detection threshold.
Abstract: Direct dark matter detection experiments based on a liquid xenon target are leading the search for dark matter particles with masses above ∼5 GeV/c2, but have limited sensitivity to lighter masses because of the small momentum transfer in dark matter-nucleus elastic scattering. However, there is an irreducible contribution from inelastic processes accompanying the elastic scattering, which leads to the excitation and ionization of the recoiling atom (the Migdal effect) or the emission of a bremsstrahlung photon. In this Letter, we report on a probe of low-mass dark matter with masses down to about 85 MeV/c2 by looking for electronic recoils induced by the Migdal effect and bremsstrahlung using data from the XENON1T experiment. Besides the approach of detecting both scintillation and ionization signals, we exploit an approach that uses ionization signals only, which allows for a lower detection threshold. This analysis significantly enhances the sensitivity of XENON1T to light dark matter previously beyond its reach.
TL;DR: In this paper, the authors present an extensive review and performance evaluation of single and multi-shot re-id algorithms, including feature extraction, metric learning, and ranking, using a large-scale dataset.
Abstract: Person re-identification (re-id) is a critical problem in video analytics applications such as security and surveillance. The public release of several datasets and code for vision algorithms has facilitated rapid progress in this area over the last few years. However, directly comparing re-id algorithms reported in the literature has become difficult since a wide variety of features, experimental protocols, and evaluation metrics are employed. In order to address this need, we present an extensive review and performance evaluation of single- and multi-shot re-id algorithms. The experimental protocol incorporates the most recent advances in both feature extraction and metric learning. To ensure a fair comparison, all of the approaches were implemented using a unified code library that includes 11 feature extraction algorithms and 22 metric learning and ranking techniques. All approaches were evaluated using a new large-scale dataset that closely mimics a real-world problem setting, in addition to 16 other publicly available datasets: VIPeR, GRID, CAVIAR, DukeMTMC4ReID, 3DPeS, PRID, V47, WARD, SAIVT-SoftBio, CUHK01, CHUK02, CUHK03, RAiD, iLIDSVID, HDA+, and Market1501. The evaluation codebase and results will be made publicly available for community use.
TL;DR: A new multi-channel gait template, called period energy image (PEI), and multi-task generative adversarial networks (MGANs), which can leverage adversarial training to extract more discriminative features from gait sequences.
Abstract: Gait recognition is of great importance in the fields of surveillance and forensics to identify human beings since gait is the unique biometric feature that can be perceived efficiently at a distance. However, the accuracy of gait recognition to some extent suffers from both the variation of view angles and the deficient gait templates. On one hand, the existing cross-view methods focus on transforming gait templates among different views, which may accumulate the transformation error in a large variation of view angles. On the other hand, a commonly used gait energy image template loses temporal information of a gait sequence. To address these problems, this paper proposes multi-task generative adversarial networks (MGANs) for learning view-specific feature representations. In order to preserve more temporal information, we also propose a new multi-channel gait template, called period energy image (PEI). Based on the assumption of view angle manifold, the MGANs can leverage adversarial training to extract more discriminative features from gait sequences. Experiments on OU-ISIR, CASIA-B, and USF benchmark data sets indicate that compared with several recently published approaches, PEI + MGANs achieves competitive performance and is more interpretable to cross-view gait recognition.
TL;DR: Deep learning-based medical image registration is a hot topic in the medical imaging research community and has achieved the state-of-the-art in many applications, including image registration as mentioned in this paper.
Abstract: The establishment of image correspondence through robust image registration is critical to many clinical tasks such as image fusion, organ atlas creation, and tumor growth monitoring, and is a very challenging problem. Since the beginning of the recent deep learning renaissance, the medical imaging research community has developed deep learning based approaches and achieved the state-of-the-art in many applications, including image registration. The rapid adoption of deep learning for image registration applications over the past few years necessitates a comprehensive summary and outlook, which is the main scope of this survey. This requires placing a focus on the different research areas as well as highlighting challenges that practitioners face. This survey, therefore, outlines the evolution of deep learning based medical image registration in the context of both research challenges and relevant innovations in the past few years. Further, this survey highlights future research directions to show how this field may be possibly moved forward to the next level.
TL;DR: To develop stable consortia, optimization of strain inoculations, nutritional divergence and crossing feeding, evolution of mutualistic growth, cell immobilization, and biosensors may potentially be used to control cell populations.
Abstract: During microbial applications, metabolic burdens can lead to a significant drop in cell performance. Novel synthetic biology tools or multi-step bioprocessing (e.g., fermentation followed by chemical conversions) are therefore needed to avoid compromised biochemical productivity from over-burdened cells. A possible solution to address metabolic burden is Division of Labor (DoL) via natural and synthetic microbial consortia. In particular, consolidated bioprocesses and metabolic cooperation for detoxification or cross feeding (e.g., vitamin C fermentation) have shown numerous successes in industrial level applications. However, distributing a metabolic pathway among proper hosts remains an engineering conundrum due to several challenges: complex subpopulation dynamics/interactions with a short time-window for stable production, suboptimal cultivation of microbial communities, proliferation of cheaters or low-producers, intermediate metabolite dilution, transport barriers between species, and breaks in metabolite channeling through biosynthesis pathways. To develop stable consortia, optimization of strain inoculations, nutritional divergence and crossing feeding, evolution of mutualistic growth, cell immobilization, and biosensors may potentially be used to control cell populations. Another opportunity is direct integration of non-bioprocesses (e.g., microbial electrosynthesis) to power cell metabolism and improve carbon efficiency. Additionally, metabolic modeling and 13C-metabolic flux analysis of mixed culture metabolism and cross-feeding offers a computational approach to complement experimental research for improved consortia performance.
TL;DR: The interference-to-noise ratio (INR) at the output of a detector is a measure of the susceptibility of a radar to interference and depends on the location of both as well as parameters such as transmit power, antenna gain, and bandwidth.
Abstract: This article examines the problem of interference in automotive radar. Different types of automotive radar as well as mechanisms and characteristics of interference and the effects of interference on radar system performance are described. The interference-to-noise ratio (INR) at the output of a detector is a measure of the susceptibility of a radar to interference. The INR is derived from different types of interfering and victim radars and depends on the location of both as well as parameters such as transmit power, antenna gain, and bandwidth. In addition, for victim radar with beamscanning, INR depends on the location of the target the victim radar is attempting to detect. Analysis is presented to show the effects of various interference scenarios on the INR. A review of the current state of the art in interference mitigation techniques previously deployed as well as areas of research currently being addressed is then provided. Finally, important future research directions are suggested.
TL;DR: In this article, the authors reviewed and synthesized the ecological impacts of road salt in freshwater ecosystems to understand species, community, and ecosystem-level responses, and identified knowledge gaps that they hope will motivate future research directions.
Abstract: Freshwater ecosystems worldwide are threatened by salinisation caused by human activities. Scientific attention on the ecological impacts of salinisation from road deicing salts is increasing exponentially. Spanning multiple trophic levels and ecosystem types, we review and synthesise the ecological impacts of road salt in freshwater ecosystems to understand species‐, community‐, and ecosystem‐level responses. In our review, we identify knowledge gaps that we hope will motivate future research directions. We found that road salts negatively affect species at all trophic levels, from biofilms to fish. The concentration at which road salt triggered an effect varied considerably. Species‐level impacts were generally sub‐lethal, leading to reductions in growth and reproduction, which can be magnified by natural stressors such as predation. Community‐level impacts including reductions of biodiversity were common, leading to communities of salt‐tolerant species, which may have implications for disease transmission from enhanced recruitment of salt‐tolerant host species such as mosquitoes. At the ecosystem level, road salts alter nutrient and energy flow. Contaminated wetlands could see greater export of greenhouse gases, streams will probably export more nitrogen and carbon, and lakes will encounter altered hydrology and oxygen dynamics, leading to greater phosphorus release from sediments. While it is necessary to keep roads safe for humans, the costs to freshwater ecosystems may be severe if actions are not taken to mitigate road salt salinisation. Cooperation among policy makers, environmental managers, transportation professionals, scientists, and the public will be crucial to prevent a loss of ecosystem services including water clarity, drinkable water, recreation venues, and fisheries.
TL;DR: In this paper, acid-catalyzed Friedel-Crafts alkylation of the polystyrene block of poly(ethylene-co-butylene)-b-polystyrene (SEBS) was used.
Abstract: Elastomeric anion exchange membranes (AEMs) were prepared by acid-catalyzed Friedel–Crafts alkylation of the polystyrene block of polystyrene-b-poly(ethylene-co-butylene)-b-polystyrene (SEBS) using...
TL;DR: It is shown that photoinduced phase separation is suppressed when perovskite nanocrystals are embedded in a non-perovskites endotaxial matrix, and the tuned bandgap remains remarkably stable under extremely intensive illumination.
Abstract: The functionality and performance of a semiconductor is determined by its bandgap. Alloying, as for instance in InxGa1-xN, has been a mainstream strategy for tuning the bandgap. Keeping the semiconductor alloys in the miscibility gap (being homogeneous), however, is non-trivial. This challenge is now being extended to halide perovskites – an emerging class of photovoltaic materials. While the bandgap can be conveniently tuned by mixing different halogen ions, as in CsPb(BrxI1-x)3, the so-called mixed-halide perovskites suffer from severe phase separation under illumination. Here, we discover that such phase separation can be highly suppressed by embedding nanocrystals of mixed-halide perovskites in an endotaxial matrix. The tuned bandgap remains remarkably stable under extremely intensive illumination. The agreement between the experiments and a nucleation model suggests that the size of the nanocrystals and the host-guest interfaces are critical for the photo-stability. The stabilized bandgap will be essential for the development of perovskite-based optoelectronics, such as tandem solar cells and full-color LEDs. The bandgap of mixed-halide perovskites can be continuously tuned by changing the halide ratio, but the crystals have poor photo-stability. Here the authors show that photoinduced phase separation is suppressed when perovskite nanocrystals are embedded in a non-perovskite endotaxial matrix.
TL;DR: This paper reviews the recent progress of chemical modifications of polysaccharides, including the common synthetic methods of chemical modification; their structural characterization; their bioactivities; and iv) the structure activity relationships of these modified poly Saccharide derivatives.
04 Feb 2019
TL;DR: In this paper, a highly reversible aqueous Zn2+ battery is demonstrated in aqueously electrolyte using V6O13·nH2O hollow microflowers composed of ultrathin nanosheets.
Abstract: Rechargeable aqueous zinc-ion batteries have been intensively studied as novel promising large-scale energy storage systems recently, owing to their advantages of high abundance, cost effectiveness, and high safety. However, the development of suitable cathode materials with superior performance is severely hampered by the sluggish kinetics of Zn2+ with divalent charge in the host structure. In the present work, a highly reversible aqueous Zn2+ battery is demonstrated in aqueous electrolyte using V6O13·nH2O hollow microflowers composed of ultrathin nanosheets. Benefiting from the synthetic merits of its favorable architecture and expanded interlamellar spacing that results from its structural water, the V6O13·nH2O cathode exhibits outstanding electrochemical performances with a high reversible capacity of 395 mAh g–1 at 0.1 A g–1, superior rate capability, and durable cycling stability with a capacity retention of 87% up to 1000 cycles. In addition, the reaction mechanism is significantly investigated in ...
TL;DR: In this article, the authors present challenges for researchers at the intersection of computer science and the social sciences in online social media, where the exploitation of this lens, termed social sensing, presents challenges.
Abstract: Online social media have democratized the broadcasting of information, encouraging users to view the world through the lens of social networks. The exploitation of this lens, termed social sensing, presents challenges for researchers at the intersection of computer science and the social sciences.
Loyola University New Orleans1, Miami University2, Australian National University3, University of Wollongong4, National Center for Atmospheric Research5, Murdoch University6, Aristotle University of Thessaloniki7, Swiss Federal Institute of Aquatic Science and Technology8, North Carolina State University9, National Institute of Water and Atmospheric Research10, Smithsonian Environmental Research Center11, University of Buenos Aires12, University of Guelph13, QIMR Berghofer Medical Research Institute14, University of Edinburgh15, Salford Royal NHS Foundation Trust16, Linnaeus University17, Rensselaer Polytechnic Institute18, University of Michigan19, University of Idaho20, Leiden University21, University of Erlangen-Nuremberg22, Finnish Meteorological Institute23, University College Cork24, University of Helsinki25, Cancer Council Victoria26, University of Gothenburg27, Columbia University28, University of Western Australia29, King's College London30, United States Environmental Protection Agency31
TL;DR: The Montreal Protocol has also played an important role in mitigating climate change as discussed by the authors, and the Montreal Protocol will continue to have far-reaching benefits for human well-being and environmental sustainability.
Abstract: Changes in stratospheric ozone and climate over the past 40-plus years have altered the solar ultraviolet (UV) radiation conditions at the Earth’s surface. Ozone depletion has also contributed to climate change across the Southern Hemisphere. These changes are interacting in complex ways to affect human health, food and water security, and ecosystem services. Many adverse effects of high UV exposure have been avoided thanks to the Montreal Protocol with its Amendments and Adjustments, which have effectively controlled the production and use of ozone-depleting substances. This international treaty has also played an important role in mitigating climate change. Climate change is modifying UV exposure and affecting how people and ecosystems respond to UV; these effects will become more pronounced in the future. The interactions between stratospheric ozone, climate and UV radiation will therefore shift over time; however, the Montreal Protocol will continue to have far-reaching benefits for human well-being and environmental sustainability.
TL;DR: This paper found that social capital, as captured by secular norms and social networks surrounding corporate headquarters, is negatively associated with levels of CEO compensation, and this relation holds in a range of robustness tests including those that address omitted variable bias and reverse causality.
01 Jul 2019
TL;DR: An abstractive meeting summarizer from both videos and audios of meeting recordings is developed, which significantly outperforms the state-of-the-art with both BLEU and ROUGE measures.
Abstract: Transcripts of natural, multi-person meetings differ significantly from documents like news articles, which can make Natural Language Generation models for generating summaries unfocused. We develop an abstractive meeting summarizer from both videos and audios of meeting recordings. Specifically, we propose a multi-modal hierarchical attention across three levels: segment, utterance and word. To narrow down the focus into topically-relevant segments, we jointly model topic segmentation and summarization. In addition to traditional text features, we introduce new multi-modal features derived from visual focus of attention, based on the assumption that the utterance is more important if the speaker receives more attention. Experiments show that our model significantly outperforms the state-of-the-art with both BLEU and ROUGE measures.