Tianjin University

Recently, channel attention mechanism has demonstrated to offer great potential in improving the performance of deep convolutional neural networks (CNNs). However, most existing methods dedicate to developing more sophisticated attention modules for achieving better performance, which inevitably increase model complexity. To overcome the paradox of performance and complexity trade-off, this paper proposes an Efficient Channel Attention (ECA) module, which only involves a handful of parameters while bringing clear performance gain. By dissecting the channel attention module in SENet, we empirically show avoiding dimensionality reduction is important for learning channel attention, and appropriate cross-channel interaction can preserve performance while significantly decreasing model complexity. Therefore, we propose a local cross-channel interaction strategy without dimensionality reduction, which can be efficiently implemented via 1D convolution. Furthermore, we develop a method to adaptively select kernel size of 1D convolution, determining coverage of local cross-channel interaction. The proposed ECA module is both efficient and effective, e.g., the parameters and computations of our modules against backbone of ResNet50 are 80 vs. 24.37M and 4.7e-4 GFlops vs. 3.86 GFlops, respectively, and the performance boost is more than 2% in terms of Top-1 accuracy. We extensively evaluate our ECA module on image classification, object detection and instance segmentation with backbones of ResNets and MobileNetV2. The experimental results show our module is more efficient while performing favorably against its counterparts.

BACKGROUND: The Global Burden of Diseases, Injuries, and Risk Factors Study 2015 provides an up-to-date synthesis of the evidence for risk factor exposure and the attributable burden of disease. By providing national and subnational assessments spanning the past 25 years, this study can inform debates on the importance of addressing risks in context. METHODS: We used the comparative risk assessment framework developed for previous iterations of the Global Burden of Disease Study to estimate attributable deaths, disability-adjusted life-years (DALYs), and trends in exposure by age group, sex, year, and geography for 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks from 1990 to 2015. This study included 388 risk-outcome pairs that met World Cancer Research Fund-defined criteria for convincing or probable evidence. We extracted relative risk and exposure estimates from randomised controlled trials, cohorts, pooled cohorts, household surveys, census data, satellite data, and other sources. We used statistical models to pool data, adjust for bias, and incorporate covariates. We developed a metric that allows comparisons of exposure across risk factors-the summary exposure value. Using the counterfactual scenario of theoretical minimum risk level, we estimated the portion of deaths and DALYs that could be attributed to a given risk. We decomposed trends in attributable burden into contributions from population growth, population age structure, risk exposure, and risk-deleted cause-specific DALY rates. We characterised risk exposure in relation to a Socio-demographic Index (SDI). FINDINGS: Between 1990 and 2015, global exposure to unsafe sanitation, household air pollution, childhood underweight, childhood stunting, and smoking each decreased by more than 25%. Global exposure for several occupational risks, high body-mass index (BMI), and drug use increased by more than 25% over the same period. All risks jointly evaluated in 2015 accounted for 57·8% (95% CI 56·6-58·8) of global deaths and 41·2% (39·8-42·8) of DALYs. In 2015, the ten largest contributors to global DALYs among Level 3 risks were high systolic blood pressure (211·8 million [192·7 million to 231·1 million] global DALYs), smoking (148·6 million [134·2 million to 163·1 million]), high fasting plasma glucose (143·1 million [125·1 million to 163·5 million]), high BMI (120·1 million [83·8 million to 158·4 million]), childhood undernutrition (113·3 million [103·9 million to 123·4 million]), ambient particulate matter (103·1 million [90·8 million to 115·1 million]), high total cholesterol (88·7 million [74·6 million to 105·7 million]), household air pollution (85·6 million [66·7 million to 106·1 million]), alcohol use (85·0 million [77·2 million to 93·0 million]), and diets high in sodium (83·0 million [49·3 million to 127·5 million]). From 1990 to 2015, attributable DALYs declined for micronutrient deficiencies, childhood undernutrition, unsafe sanitation and water, and household air pollution; reductions in risk-deleted DALY rates rather than reductions in exposure drove these declines. Rising exposure contributed to notable increases in attributable DALYs from high BMI, high fasting plasma glucose, occupational carcinogens, and drug use. Environmental risks and childhood undernutrition declined steadily with SDI; low physical activity, high BMI, and high fasting plasma glucose increased with SDI. In 119 countries, metabolic risks, such as high BMI and fasting plasma glucose, contributed the most attributable DALYs in 2015. Regionally, smoking still ranked among the leading five risk factors for attributable DALYs in 109 countries; childhood underweight and unsafe sex remained primary drivers of early death and disability in much of sub-Saharan Africa. INTERPRETATION: Declines in some key environmental risks have contributed to declines in critical infectious diseases. Some risks appear to be invariant to SDI. Increasing risks, including high BMI, high fasting plasma glucose, drug use, and some occupational exposures, contribute to rising burden from some conditions, but also provide opportunities for intervention. Some highly preventable risks, such as smoking, remain major causes of attributable DALYs, even as exposure is declining. Public policy makers need to pay attention to the risks that are increasingly major contributors to global burden. FUNDING: Bill & Melinda Gates Foundation.

A fundamental change has been achieved in understanding surface electrochemistry due to the profound knowledge of the nature of electrocatalytic processes accumulated over the past several decades and to the recent technological advances in spectroscopy and high resolution imaging. Nowadays one can preferably design electrocatalysts based on the deep theoretical knowledge of electronic structures, via computer-guided engineering of the surface and (electro)chemical properties of materials, followed by the synthesis of practical materials with high performance for specific reactions. This review provides insights into both theoretical and experimental electrochemistry toward a better understanding of a series of key clean energy conversion reactions including oxygen reduction reaction (ORR), oxygen evolution reaction (OER), and hydrogen evolution reaction (HER). The emphasis of this review is on the origin of the electrocatalytic activity of nanostructured catalysts toward the aforementioned reactions by correlating the apparent electrode performance with their intrinsic electrochemical properties. Also, a rational design of electrocatalysts is proposed starting from the most fundamental aspects of the electronic structure engineering to a more practical level of nanotechnological fabrication.

Bounding box regression is the crucial step in object detection. In existing methods, while ℓn-norm loss is widely adopted for bounding box regression, it is not tailored to the evaluation metric, i.e., Intersection over Union (IoU). Recently, IoU loss and generalized IoU (GIoU) loss have been proposed to benefit the IoU metric, but still suffer from the problems of slow convergence and inaccurate regression. In this paper, we propose a Distance-IoU (DIoU) loss by incorporating the normalized distance between the predicted box and the target box, which converges much faster in training than IoU and GIoU losses. Furthermore, this paper summarizes three geometric factors in bounding box regression, i.e., overlap area, central point distance and aspect ratio, based on which a Complete IoU (CIoU) loss is proposed, thereby leading to faster convergence and better performance. By incorporating DIoU and CIoU losses into state-of-the-art object detection algorithms, e.g., YOLO v3, SSD and Faster R-CNN, we achieve notable performance gains in terms of not only IoU metric but also GIoU metric. Moreover, DIoU can be easily adopted into non-maximum suppression (NMS) to act as the criterion, further boosting performance improvement. The source code and trained models are available at https://github.com/Zzh-tju/DIoU.

Covalent organic frameworks (COFs) are a class of crystalline porous organic polymers with permanent porosity and highly ordered structures. Unlike other polymers, a significant feature of COFs is that they are structurally predesignable, synthetically controllable, and functionally manageable. In principle, the topological design diagram offers geometric guidance for the structural tiling of extended porous polygons, and the polycondensation reactions provide synthetic ways to construct the predesigned primary and high-order structures. Progress over the past decade in the chemistry of these two aspects undoubtedly established the base of the COF field. By virtue of the availability of organic units and the diversity of topologies and linkages, COFs have emerged as a new field of organic materials that offer a powerful molecular platform for complex structural design and tailor-made functional development. Here we target a comprehensive review of the COF field, provide a historic overview of the chemistry of the COF field, survey the advances in the topology design and synthetic reactions, illustrate the structural features and diversities, scrutinize the development and potential of various functions through elucidating structure-function correlations based on interactions with photons, electrons, holes, spins, ions, and molecules, discuss the key fundamental and challenging issues that need to be addressed, and predict the future directions from chemistry, physics, and materials perspectives.

Electrolytes have been identified as some of the most influential components in the performance of electrochemical supercapacitors (ESs), which include: electrical double-layer capacitors, pseudocapacitors and hybrid supercapacitors. This paper reviews recent progress in the research and development of ES electrolytes. The electrolytes are classified into several categories, including: aqueous, organic, ionic liquids, solid-state or quasi-solid-state, as well as redox-active electrolytes. Effects of electrolyte properties on ES performance are discussed in detail. The principles and methods of designing and optimizing electrolytes for ES performance and application are highlighted through a comprehensive analysis of the literature. Interaction among the electrolytes, electro-active materials and inactive components (current collectors, binders, and separators) is discussed. The challenges in producing high-performing electrolytes are analyzed. Several possible research directions to overcome these challenges are proposed for future efforts, with the main aim of improving ESs' energy density without sacrificing existing advantages (e.g., a high power density and a long cycle-life) (507 references).

Owing to the increasing emissions of carbon dioxide (CO(2)), human life and the ecological environment have been affected by global warming and climate changes. To mitigate the concentration of CO(2) in the atmosphere various strategies have been implemented such as separation, storage, and utilization of CO(2). Although it has been explored for many years, hydrogenation reaction, an important representative among chemical conversions of CO(2), offers challenging opportunities for sustainable development in energy and the environment. Indeed, the hydrogenation of CO(2) not only reduces the increasing CO(2) buildup but also produces fuels and chemicals. In this critical review we discuss recent developments in this area, with emphases on catalytic reactivity, reactor innovation, and reaction mechanism. We also provide an overview regarding the challenges and opportunities for future research in the field (319 references).

The urgent need of clean and renewable energy drives the exploration of effective strategies to produce molecular hydrogen. With the assistance of highly active non-noble metal electrocatalysts, electrolysis of water is becoming a promising candidate to generate pure hydrogen with low cost and high efficiency. Very recently, transition metal phosphides (TMPs) have been proven to be high performance catalysts with high activity, high stability, and nearly ∼100% Faradic efficiency in not only strong acidic solutions, but also in strong alkaline and neutral media for electrochemical hydrogen evolution. In this tutorial review, an overview of recent development of TMP nanomaterials as catalysts for hydrogen generation with high activity and stability is presented. The effects of phosphorus (P) on HER activity, and their synthetic methods of TMPs are briefly discussed. Then we will demonstrate the specific strategies to further improve the catalytic efficiency and stability of TMPs by structural engineering. Making use of TMPs as cocatalysts and catalysts in photochemical and photoelectrochemical water splitting is also discussed. Finally, some key challenges and issues which should not be ignored during the rapid development of TMPs are pointed out. These strategies and challenges of TMPs are instructive for designing other high-performance non-noble metal catalysts.

Electricity-driven water splitting can facilitate the storage of electrical energy in the form of hydrogen gas. As a half-reaction of electricity-driven water splitting, the oxygen evolution reaction (OER) is the major bottleneck due to the sluggish kinetics of this four-electron transfer reaction. Developing low-cost and robust OER catalysts is critical to solving this efficiency problem in water splitting. The catalyst design has to be built based on the fundamental understanding of the OER mechanism and the origin of the reaction overpotential. In this article, we summarize the recent progress in understanding OER mechanisms, which include the conventional adsorbate evolution mechanism (AEM) and lattice-oxygen-mediated mechanism (LOM) from both theoretical and experimental aspects. We start with the discussion on the AEM and its linked scaling relations among various reaction intermediates. The strategies to reduce overpotential based on the AEM and its derived descriptors are then introduced. To further reduce the OER overpotential, it is necessary to break the scaling relation of HOO* and HO* intermediates in conventional AEM to go beyond the activity limitation of the volcano relationship. Strategies such as stabilization of HOO*, proton acceptor functionality, and switching the OER pathway to LOM are discussed. The remaining questions on the OER and related perspectives are also presented at the end.

Graphene materials (GMs) as supercapacitor electrode materials have been investigated. GMs are prepared from graphene oxide sheets, and subsequently suffer a gas-based hydrazine reduction to restore the conducting carbon network. A maximum specific capacitance of 205 F/g with a measured power density of 10 kW/kg at energy density of 28.5 Wh/kg in an aqueous electrolyte solution has been obtained. Meanwhile, the supercapacitor devices exhibit excellent long cycle life along with ∼90% specific capacitance retained after 1200 cycle tests. These remarkable results demonstrate the exciting commercial potential for high performance, environmentally friendly and low-cost electrical energy storage devices based on this new 2D graphene material.

Earth-abundant first-row (3d) transition metal-based catalysts have been developed for the oxygen-evolution reaction (OER); however, they operate at overpotentials substantially above thermodynamic requirements. Density functional theory suggested that non-3d high-valency metals such as tungsten can modulate 3d metal oxides, providing near-optimal adsorption energies for OER intermediates. We developed a room-temperature synthesis to produce gelled oxyhydroxides materials with an atomically homogeneous metal distribution. These gelled FeCoW oxyhydroxides exhibit the lowest overpotential (191 millivolts) reported at 10 milliamperes per square centimeter in alkaline electrolyte. The catalyst shows no evidence of degradation after more than 500 hours of operation. X-ray absorption and computational studies reveal a synergistic interplay between tungsten, iron, and cobalt in producing a favorable local coordination environment and electronic structure that enhance the energetics for OER.

A very basic pathway from CO 2 to ethylene Ethylene is an important commodity chemical for plastics. It is considered a tractable target for synthesizing renewably from carbon dioxide (CO 2 ). The challenge is that the performance of the copper electrocatalysts used for this conversion under the required basic reaction conditions suffers from the competing reaction of CO 2 with the base to form bicarbonate. Dinh et al. designed an electrode that tolerates the base by optimizing CO 2 diffusion to the catalytic sites (see the Perspective by Ager and Lapkin). This catalyst design delivers 70% efficiency for 150 hours. Science , this issue p. 783 ; see also p. 707

The paper presents a novel method, Zero-Reference Deep Curve Estimation (Zero-DCE), which formulates light enhancement as a task of image-specific curve estimation with a deep network. Our method trains a lightweight deep network, DCE-Net, to estimate pixel-wise and high-order curves for dynamic range adjustment of a given image. The curve estimation is specially designed, considering pixel value range, monotonicity, and differentiability. Zero-DCE is appealing in its relaxed assumption on reference images, i.e., it does not require any paired or unpaired data during training. This is achieved through a set of carefully formulated non-reference loss functions, which implicitly measure the enhancement quality and drive the learning of the network. Our method is efficient as image enhancement can be achieved by an intuitive and simple nonlinear curve mapping. Despite its simplicity, we show that it generalizes well to diverse lighting conditions. Extensive experiments on various benchmarks demonstrate the advantages of our method over state-of-the-art methods qualitatively and quantitatively. Furthermore, the potential benefits of our Zero-DCE to face detection in the dark are discussed.

This review describes the current understanding of CO₂ photoreduction on the surface of heterogeneous catalysts with a particular focus on the reaction mechanism and pathways as well as the adsorption/activation of CO₂.

This study proposes a novel unified and unsupervised end-to-end image fusion network, termed as U2Fusion, which is capable of solving different fusion problems, including multi-modal, multi-exposure, and multi-focus cases. Using feature extraction and information measurement, U2Fusion automatically estimates the importance of corresponding source images and comes up with adaptive information preservation degrees. Hence, different fusion tasks are unified in the same framework. Based on the adaptive degrees, a network is trained to preserve the adaptive similarity between the fusion result and source images. Therefore, the stumbling blocks in applying deep learning for image fusion, e.g., the requirement of ground-truth and specifically designed metrics, are greatly mitigated. By avoiding the loss of previous fusion capabilities when training a single model for different tasks sequentially, we obtain a unified model that is applicable to multiple fusion tasks. Moreover, a new aligned infrared and visible image dataset, RoadScene (available at https://github.com/hanna-xu/RoadScene), is released to provide a new option for benchmark evaluation. Qualitative and quantitative experimental results on three typical image fusion tasks validate the effectiveness and universality of U2Fusion. Our code is publicly available at https://github.com/hanna-xu/U2Fusion.

Stable graphene suspensions are prepared using a new, green, and intriguing synthetic route. Graphene can be directly prepared by the fast deoxygenation of exfoliated graphite oxide in strong alkaline solutions at low temperatures (50–90 °C), in the absence of reducing agents. This study opens an opportunity towards the production of processable graphene on an industrial scale. Detailed facts of importance to specialist readers are published as ”Supporting Information”. Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

A novel coronavirus [severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2)] outbreak has caused a global coronavirus disease 2019 (COVID-19) pandemic, resulting in tens of thousands of infections and thousands of deaths worldwide. The RNA-dependent RNA polymerase [(RdRp), also named nsp12] is the central component of coronaviral replication and transcription machinery, and it appears to be a primary target for the antiviral drug remdesivir. We report the cryo-electron microscopy structure of COVID-19 virus full-length nsp12 in complex with cofactors nsp7 and nsp8 at 2.9-angstrom resolution. In addition to the conserved architecture of the polymerase core of the viral polymerase family, nsp12 possesses a newly identified β-hairpin domain at its N terminus. A comparative analysis model shows how remdesivir binds to this polymerase. The structure provides a basis for the design of new antiviral therapeutics that target viral RdRp.

Alkaline water splitting is an attractive method for sustainable hydrogen production.

Due to limitations of low power density, high cost, heavy weight, etc., the development and application of battery-powered devices are facing with unprecedented technical challenges. As a novel pattern of energization, the wireless power transfer (WPT) offers a band new way to the energy acquisition for electric-driven devices, thus alleviating the over-dependence on the battery. This paper presents an overview of WPT techniques with emphasis on working mechanisms, technical challenges, metamaterials, and classical applications. Focusing on WPT systems, this paper elaborates on current major research topics and discusses about future development trends. This novel energy transmission mechanism shows significant meanings on the pervasive application of renewable energies in our daily life.

Hydrogen evolution reaction (HER) in alkaline medium is currently a point of focus for sustainable development of hydrogen as an alternative clean fuel for various energy systems, but suffers from sluggish reaction kinetics due to additional water dissociation step. So, the state-of-the-art catalysts performing well in acidic media lose considerable catalytic performance in alkaline media. This review summarizes the recent developments to overcome the kinetics issues of alkaline HER, synthesis of materials with modified morphologies, and electronic structures to tune the active sites and their applications as efficient catalysts for HER. It first explains the fundamentals and electrochemistry of HER and then outlines the requirements for an efficient and stable catalyst in alkaline medium. The challenges with alkaline HER and limitation with the electrocatalysts along with prospective solutions are then highlighted. It further describes the synthesis methods of advanced nanostructures based on carbon, noble, and inexpensive metals and their heterogeneous structures. These heterogeneous structures provide some ideal systems for analyzing the role of structure and synergy on alkaline HER catalysis. At the end, it provides the concluding remarks and future perspectives that can be helpful for tuning the catalysts active-sites with improved electrochemical efficiencies in future.