Search

Abstract Background Extensive evidence indicates that both lifestyle factors and air pollution are strongly associated with all-cause mortality. However, little studies in this field have integrated these two factors in order to examine their relationship with mortality and explore potential interactions. Methods A cohort of 271,075 participants from the UK Biobank underwent analysis. Lifestyles in terms of five modifiable factors, namely smoking, alcohol consumption, physical activity, diet, and sleep quality, were classified as unhealthy (0–1 score), general (2–3 score), and healthy (4–5 score). Air pollution, including particle matter with a diameter ≤ 2.5 μm (PM2.5), particulate matter with a diameter ≤ 10 μm (PM10), particulate matter with a diameter 2.5–10 μm (PM2.5−10), nitrogen dioxide (NO2), and nitrogen oxides (NOx), was divided into three levels (high, moderate, and low) using Latent Profile Analysis (LPA). Cox proportional hazard regression analysis was performed to examine the links between lifestyle, air pollution, and all-cause mortality before and after adjustment for potential confounders. Restricted cubic spline curves featuring three knots were incorporated to determine nonlinear relationships. The robustness of the findings was assessed via subgroup and sensitivity analyses. Results With unhealthy lifestyles have a significantly enhanced risk of death compared to people with general lifestyles (HR = 1.315, 95% CI, 1.277–1.355), while people with healthy lifestyles have a significantly lower risk of death (HR = 0.821, 95% CI, 0.785–0.858). Notably, the difference in risk between moderate air pollution and mortality risk remained insignificant (HR = 0.993, 95% CI, 0.945–1.044). High air pollution, on the other hand, was independently linked to increased mortality risk as compared to low air pollution (HR = 1.162, 95% CI, 1.124–1.201). The relationship between NOx, PM10, and PM2.5−10 and all-cause mortality was found to be nonlinear (p for nonlinearity

In data acquisition and transmission, hyperspectral images are inevitably corrupted by additive noises, making it challenging to accurately observe and recognize the materials on the surface of the Earth. However, scholars have been addicted to developing numerous complex methods for separable two-stage denoising and anomaly detection (AD) tasks over the past years, rarely paying attention to the real effect of noises for subsequent intelligent interpretation. To this date, we propose a hierarchical integration framework for hyperspectral simultaneous AD and denoising (HyADD). Joint AD and denoising are mutually integrated and their outputs in each iteration stimulate each other, breaking through the respective performance bottlenecks of the separable two-stage scheme. Inspired by spatial–spectral gradient domain-based constraint, HyADD removes additive noises and preserves advantageous image smoothness information to improve intermediate detection performances in the iteration loop. Conversely, with the assistance of the antinoise dictionary conduction and the subspace domain-based low-rankness, the identification of anomalies with different features from the background can provide effective feedback to the denoising process. The proposed algorithm is efficiently solved by the well-designed linearized alternating direction method of multipliers with an adaptive penalty. A comparison with the existing well-known AD algorithms via simulated and real-world experiments establishes the competitiveness of the proposed HyADD with state-of-the-art methods.

Extraction of vectorized building instances from remote sensing images has made significant progress in recent years. The extraction of vectorized building maps enables rapid, large-scale updates to geospatial databases. Recently, the contour-based methods have achieved amazing success because their learning goals are more compatible with vectorized polygons. However, existing contour-based approaches achieve contour updates only by learning the local features at each vertex, which can lead to a contour smoothing problem. Besides, these methods achieve classification via features at the center point, which hinders the receptive field at the instance level. Our primary motivation is to overcome existing limitations introduced by local modeling during contour regression while simultaneously improving the instance classification performance. In this article, we utilize instance-level features to guide the contour learning process. An instance-aware contour regression (IACR) module is designed to update the local contour features via cross attention on the instance-level features. Furthermore, based on the IACR module, we propose a novel vectorized building extraction framework BuildingVec, which interacts between the contour regression branch and instance classification branch through a well-designed cascade architecture. We also build a vectorized building dataset with fine-grained categories, named UBCv2 (vec), to benefit the study of vectorized extraction for fine-grained buildings. Experiments on the UBCv2 (vec) dataset demonstrate that BuildingVec achieves state-of-the-art performance compared to both mask-based methods and contour-based methods. When compared to other vectorized extraction methods on the Crowd AI dataset, BuildingVec also achieves a state-of-the-art performance with an $\text{AP}_{50}$ of 93.1%.

This article presents the scientific outcomes of the 2023 Data Fusion Contest (DFC23) organized by the Image Analysis and Data Fusion Technical Committee of the IEEE Geoscience and Remote Sensing Society. The contest consists of two tracks investigating the fusion of optical and synthetic aperture radar data for: 1) fine-grained roof type classification and 2) height estimation. During the development phase, 1000 people registered for the contest, while at the end 55 and 35 teams competed during the test phase in the two tracks, respectively. This article presents the methods and results obtained by the first and second-ranked teams of each track. In Track 1, both winning teams leveraged pretraining, modern network architectures, model ensembles, and measures to cope with the imbalanced class distribution. The solutions to Track 2 are more diverse and are characterized by modern multitask learning approaches. The data of this contest is openly available to the community for further research, development, and refinement of machine learning methods.

Background: Eosinophilic esophagitis (EoE) is a chronic, inflammatory disease of the esophagus leading to symptoms of esophageal dysfunction; dysphagia is the most common symptom experienced by adults and adolescents. Objective: We sought to perform a psychometric evaluation of the Dysphagia Symptom Questionnaire (DSQ), a patient-reported outcome measure for patients with EoE. Methods: Using baseline and week 24 data from the randomized, interventional, multinational phase 3 R668-EE-1774 trial (NCT03633617), the measurement properties of the DSQ—including reliability, construct and known-groups validity, responsiveness, and interpretation of change—were evaluated. Results: The analysis population comprised 239 patients with EoE (age [mean ± SD], 28.1 ± 13.14 years; 63.6% male; 90.4% White). Intraclass correlation coefficients of 0.92 and 0.97 exceeded the acceptable reliability threshold (≥0.70). Construct validity correlations with EoE symptom and impact measures were moderate at baseline (|r| = 0.44-0.55) and week 24 (|r| = 0.55-0.69), and the DSQ biweekly total score discriminated among groups defined by disease severity. Analyses exploring interpretation of change from baseline on the DSQ biweekly total score indicated thresholds for within-patient improvement ranging from 9 to 23 points; a within-patient improvement from baseline of 13 points or greater could be considered clinically meaningful. Conclusions: This analysis confirmed that the DSQ has acceptable distributional properties, test-retest reliability, construct validity, and ability to detect change. Therefore, the DSQ is a valid and reliable measure to assess the patient-reported symptom of dysphagia among adult and adolescent patients with EoE in the context of a clinical trial setting.

Abstract Background Eosinophilic esophagitis (EoE) has a detrimental effect on health-related quality of life (HRQOL). The Eosinophilic Esophagitis Impact Questionnaire (EoE-IQ) is a novel patient-reported outcome (PRO) measure assessing the impact of EoE on HRQOL. To assess suitability of the EoE-IQ, its measurement properties were evaluated. Methods Using baseline and week 24 data from the pivotal, randomized, placebo-controlled, multinational phase 3 R668-EE-1774 trial (NCT03633617) of dupilumab, we evaluated EoE-IQ’s measurement properties (including reliability, construct and known-groups validity, and ability to detect change) and established the threshold for change in scores that can be considered clinically meaningful. Results The analysis population comprised 239 adults and adolescents with EoE. Mean age was 28.1 (standard deviation, 13.14) years; 63.6% were male, and 90.4% were White. Reliability estimates for the EoE-IQ average score exceeded acceptable thresholds for patients who were stable as indicated by ratings of Patient Global Impression of Severity (PGIS) and Change (PGIC) (intraclass correlation coefficients, 0.75 and 0.81). Construct validity correlations with other EoE-specific PRO scores were moderate at baseline (|r|= 0.44–0.60) and moderate to strong at week 24 (|r|= 0.61–0.72). In known-groups analysis, EoE-IQ average score discriminated among groups of patients at varying EoE severity levels defined by PGIS scores. A ≥ 0.6-point reduction in EoE-IQ average score (where scores range from 1 to 5, with higher scores indicating worse HRQOL) from baseline to week 24 can be considered clinically meaningful. Conclusions The EoE-IQ’s measurement properties are acceptable, making it a valid, reliable measure of the HRQOL impacts of EoE among adults and adolescents. Trial registration: ClinicalTrials.gov, NCT03633617. Registered August 14, 2018, https://clinicaltrials.gov/study/NCT03633617 .

Abstract Background Many studies exist on the living arrangements and health status of older adults, but the findings have been inconsistent. Therefore, we examined the relationship between living arrangements and all-cause mortality in older adults. Methods This perspective study was based on the Chinese Longitudinal Healthy Longevity Survey (CLHLS) from 2011 to 2018. We used a sample aged 65 years and over included in the study in 2011. Propensity score matching was performed to minimize bias and Cox proportional hazards regression models were conducted. Results A total of 7,963 participants were included. Of these, 1,383 were living alone, 6,424 were living with families, and 156 were living in nursing homes. In the propensity score-matched cohort, older adults living alone had a significantly lower risk of all-cause mortality than those living with families (hazard ratio 0.85; 95% confidence intervals 0.76 to 0.95). Living alone was prominently associated with a decline in mortality compared with living in nursing homes (hazard ratio 0.61; 95% confidence intervals 0.44 to 0.84). There was no significant difference in mortality between living in nursing homes and living with families (hazard ratio 1.19; 95% confidence intervals 0.89 to 1.60). Subgroup analyses indicated that there was no significant interaction with age, sex, education, or residence. Conclusions The risk of all-cause mortality was significantly lower in older adults living alone than in those living with families or living in nursing homes. This article’s findings suggest the need to adopt multiple approaches to meet the needs of senior care services.

Porcine deltacoronavirus (PDCoV) is an emerging enteric pathogen that causes substantial economic losses in the swine industry worldwide. The PDCoV NS6 protein is an accessory protein that plays a pivotal role in the viral life cycle and immune evasion. However, the functions of NS6 and its role in PDCoV pathogenesis remain largely unknown. In this study, we prepared a monoclonal antibody (mAb) 5-A11 that specifically recognizes the PDCoV NS6 protein. The mAb 5-A11 exhibited high specificity for PDCoV, with no cross-reactivity with several major porcine pathogenic viruses. Furthermore, the epitope recognized by mAb 5-A11 was precisely mapped to residues 70EYGSIYGKDFI80 of the NS6 protein using Western blot analysis. Notably, this epitope is highly conserved among different PDCoV isolates. Substantial variations were observed when comparing this epitope with the corresponding regions in the NS6 proteins of other δ coronaviruses, suggesting potential differences in the structure, function, and antigenicity of their NS6 proteins. Our findings provide valuable tools and insights for further elucidating the functions of the NS6 protein and its role in PDCoV pathogenesis, as well as for developing diagnostic and therapeutic strategies against PDCoV infection.

Collaborative perception enhances onboard perceptual capability by integrating features from other platforms, effectively mitigating the compromised accuracy caused by a restricted observational range and vulnerability to interference. However, current implementations of collaborative perception overlook the prevalent issues of both limited and low-reliability communication, as well as misaligned observations in remote sensing. To address this problem, this article presents an innovative distributed collaborative perception network (DCP-Net) specifically designed for remote sensing applications. Firstly, a self-mutual information match module is proposed to identify collaboration opportunities and select suitable partners. This module prioritizes critical collaborative features and reduces redundant transmission for better adaptation to weak communication in remote sensing. Secondly, a related feature fusion module is devised to tackle the misalignment between local and collaborative features due to the multiangle observations, improving the quality of fused features for the downstream task. We conduct extensive experiments and visualization analyses using three semantic segmentation datasets, namely Potsdam, iSAID, and DFC23. The results demonstrate that DCP-Net outperforms the existing collaborative perception methods comprehensively, improving mIoU by 2.61% to 16.89% at the highest collaboration efficiency and achieving state-of-the-art performance.

Semantic segmentation with cross-domain adaptation in remote-sensing images (RSIs) is crucial and mitigates the expense of manually labeling target data. However, the performance of existing unsupervised domain adaptation (UDA) methods is still significantly impacted by domain bias, leading to a considerable gap compared to supervised trained models. To address this, our work focuses on semi-supervised domain adaptation, selecting a small subset of target annotations through active learning (AL) that maximize information to improve domain adaptation. Overall, we propose a novel active bidirectional self-training network (ABSNet) for cross-domain semantic segmentation in RSIs. ABSNet consists of two sub-stages: a multi-prototype active region selection (MARS) stage and a source-weighted class-balanced self-training (SCBS) stage. The MARS approach captures the diversity in labeled source data by introducing multi-prototype density estimation based on Gaussian mixture models. We then measure inter-domain similarity to select complementary and representative target samples. Through fine-tuning with the selected active samples, we propose an enhanced self-training strategy SCBS, designed for weighted training on source data, aiming to avoid the negative effects of interfering samples. We conduct extensive experiments on the LoveDA and ISPRS datasets to validate the superiority of our method over existing state-of-the-art domain-adaptive semantic segmentation methods.

ABSTRACTSpecies of the phylum Actinobacteria have long time been thought mostly non-motile and non-flagellated, in spite of their extraordinary diversity in morphology, physiology, and ecology. It remains unclear how very few actinobacterial species acquired their flagellar genes and the role of flagellar motility in the evolution of different lineages within this large phylum. Here, we performed a comprehensive phylogenomic analysis of flagellar components in all actinobacterial species, including the deepest branches of this phylum identified recently. Our results provide robust evidence that all actinobacterial flagella shared a common origin and likely evolved from their last common flagellated ancestor by vertical inheritance. Later lineages underwent massive flagellar losses with the ecological switch from aquatic to terrestrial land or host-associated environments. In addition, the actinobacterial flagella also showed degenerative evolution in terms of their composition, leading to the simplest rod known so far. Furthermore, these evolutionary changes of flagella were accompanied by noteworthy variations in the chemosensory system and c-di-GMP mediated signaling network. Overall, we present here a detailed picture of flagellar evolution in Actinobacteria, which provides new insights into the evolution of major actinobacterial lineages and previously unobserved links between flagella and other biological features. The simplest flagellar rod as a product of degenerative evolution could possibly serve as a model for future nanomachine reconstruction in synthetic biology.IMPORTANCEFlagellar motility plays an important role in the environmental adaptation of bacteria and is found in more than 50% of known bacterial species. However, this important characteristic is sparsely distributed within members of the phylum Actinobacteria, which constitutes one of the largest bacterial groups. It is unclear why this important fitness organelle is absent in most actinobacterial species and the origin of flagellar genes in other species. Here, we present detailed analyses of the evolution of flagellar genes in Actinobacteria, in conjunction with the ecological distribution and cell biological features of major actinobacterial lineages, and the co-evolution of signal transduction systems. The results presented in addition to clarifying the puzzle of sporadic distribution of flagellar motility in Actinobacteria, also provide important insights into the evolution of major lineages within this phylum.

This study proposes a Synthetic Aperture Radar (SAR) aircraft detection and recognition method combined with scattering perception to address the problem of target discreteness and false alarms caused by strong background interference in SAR images. The global information is enhanced through a context-guided feature pyramid module, which suppresses strong disturbances in complex images and improves the accuracy of detection and recognition. Additionally, scatter key points are used to locate targets, and a scatter-aware detection module is designed to realize the fine correction of the regression boxes to improve target localization accuracy. This study generates and presents a high-resolution SAR-AIRcraft-1.0 dataset to verify the effectiveness of the proposed method and promote the research on SAR aircraft detection and recognition. The images in this dataset are obtained from the satellite Gaofen-3, which contains 4,368 images and 16,463 aircraft instances, covering seven aircraft categories, namely A220, A320/321, A330, ARJ21, Boeing737, Boeing787, and other. We apply the proposed method and common deep learning algorithms to the constructed dataset. The experimental results demonstrate the excellent effectiveness of our method combined with scattering perception. Furthermore, we establish benchmarks for the performance indicators of the dataset in different tasks such as SAR aircraft detection, recognition, and integrated detection and recognition.

Synthetic aperture radar (SAR) ship detection and classification has gained unprecedented attention due to its important role in maritime transportation. Many deep learning-based detectors and classifiers have been successfully applied and achieved great progress. However, ships in SAR images present discrete and multi-centric features, and their scattering characteristics and edge information are sensitive to variations in target attitude angles (TAAs). These factors pose challenges for existing methods to obtain satisfactory results. To address these challenges, a novel target attitude angle-guided network (TAG-Net) is proposed in this article. The core idea of TAG-Net is to leverage TAA information as guidance and use an adaptive feature-level fusion strategy to dynamically learn more representative features that can handle the target imaging diversity caused by TAA. This is achieved through a TAA-aware feature modulation (TAFM) module. It uses the TAA information and foreground information as prior knowledge and establishes the relationship between the ship scattering characteristics and TAA information. This enables a reduction in the intra-class variability and highlights ship targets. Additionally, considering the different requirements of the detection and classification tasks for the scattering information, we propose a layer-wise attention-based task decoupling detection head (LATD). Unlike general deep learning methods that use shared features for both detection and classification tasks, LATD extracts multi-level features and uses layer attention to achieve feature decoupling and select the most suitable features for each task. Finally, we introduce a novel salient-enhanced feature balance module (SFB) to provide richer semantic information and capture the global context to highlight ships in complex scenes, effectively reducing the impact of background noise. A large-scale ship detection dataset (LSSDD+) is used to verify the effectiveness of TAG-Net, and our method achieves state-of-the-art performance.

The effect of rare earth metals added to wire on droplet transfer mode in GMAW with pure CO2 shielding gas was reviewed. The results show that when the welding wire containing rare earth metals is welded with DCEN pure CO2 shielding gas, owing to the expansion of the cathode spot area in the arc (the arc above the droplet presents a conical shape), the refined droplets form an axial spray transfer mode, the spatter generation is very small, and the usability is excellent. The rare earth metals reduce the work function of the cathode electrons, at the same time, the high electromagnetic pinch force allows the droplets to transition axially in a smaller size, and the direction of the dominant force acting on the droplets facilitates the droplet transfer, and finally the formation conditions of droplet spray transfer are fully satisfied under pure CO2 shielding gas. The application of welding wire containing rare earth metal in new power supply with current waveform control makes the J-STAR® welding method realize ultra-low spatter in a wide current range, which widens the application range of the process.

Dumpsites are hard to locate globally. Here the authors apply deep networks to satellite images to provide an effective and low-cost way to detect dumpsites with the new method saving more than 96.8% of the manual time with a strong sensitivity to dumpsites.

Immunotherapy, in the form of hematopoietic stem cell transplantation (HSCT), has been part of the standard of care in the treatment of acute leukemia for over 40 years. Trials evaluating novel immunotherapeutic approaches, such as targeting the programmed death-1 (PD-1) pathway, have unfortunately not yielded comparable results to those seen in solid tumors. Major histocompatibility complex (MHC) proteins are cell surface proteins essential for the adaptive immune system to recognize self versus non-self. MHC typing is used to determine donor compatibility when evaluating patients for HSCT. Recently, loss of MHC class II (MHC II) was shown to be a mechanism of immune escape in patients with acute myeloid leukemia after HSCT. Here we report that treatment with the tyrosine kinase inhibitor, dasatinib, and an anti-PD-1 antibody in preclinical models of Philadelphia chromosome positive B-cell acute lymphoblastic leukemia is highly active. The dasatinib and anti-PD-1 combination reduces tumor burden, is efficacious, and extends survival. Mechanistically, we found that treatment with dasatinib significantly increased MHC II expression on the surface of antigen-presenting cells (APC) in a tumor microenvironment-independent fashion and caused influx of APC cells into the leukemic bone marrow. Finally, the induction of MHC II may potentiate immune memory by impairing leukemic engraftment in mice previously cured with dasatinib, after re-inoculation of leukemia cells. In summary, our data suggests that anti-PD-1 therapy may enhance the killing ability of dasatinib via dasatinib driven APC growth and expansion and upregulation of MHC II expression, leading to antileukemic immune rewiring.

Detection of damaged buildings is a form of object detection and is essential for disaster emergency response efforts. In recent years, deep learning has been widely used in object detection, with successful target detection models such as Faster-Rcnn and You Only Look Once (YOLO) being proposed. However, training deep learning models usually requires a large amount of labeled data. Due to the high threshold for aerial remote sensing data collection, labeled aerial data of collapsed buildings is very sparse. In addition, the limited area of damage in a single scene leads to insufficient feature diversity, which can easily lead to model overfitting. These issues restrict the development of deep learning in emergency response applications. To solve these problems, we propose a paradigm named cross-domain teacher–student mutual training. By using the Cycle-GAN-generated style transfer data through teacher network, pseudolabels are generated to train the student network. Then, the student network slowly updates the parameters of the teacher network to indirectly learn the generalization information of the satellite data domain. Networks trained in this way can achieve good results in detecting collapsed houses in aviation and satellite data. We tested the results on our self-built dataset, DB-ARSD, which includes bounding box labeling of the damaged buildings, and found that our method outperforms other object detection methods in both collapsed house prediction accuracy and domain transfer generalization performance.

With the growth of remote sensing images, understanding image content automatically has attracted many researchers' interests in deep learning for remote sensing image. Inspired from the natural image captioning, the model with convolutional neural network (CNN)-Recurrent neural network (RNN) as the backbone and supplemented by attention has been widely used in remote sensing image captioning. However, it is inefficient for the current attention layer to simultaneously mine hidden foreground from the background of remote sensing image and perform feature interactive learning. Meanwhile, the new mainstream language model has recently surpassed the traditional long short-term memory (LSTM) in sentence generation. For solving the above problems, in this article, we proposed a novel thought to make the flat remote sensing images stereoscopic by separating the foreground and background. Based on hierarchical image information, we designed a novel Deformable Transformer equipped with deformable scaled dot-product attention to learn multiscale feature from foreground and background through the powerful interactive learning ability. Evaluations are conducted on four classic remote sensing image captioning datasets. Compared with the state-of-the-art methods, our Transformer variant achieves higher captioning accuracy.

Intelligent processing of satellite video focuses on extracting specific information of ground objects and scenes from earth observation videos through intelligent image/video processing technology, which has important applications in fields such as traffic monitoring, resource monitoring, and environmental monitoring. The integration of deep learning technology in satellite video processing has led to significant advancements in tasks such as object detection and object tracking, expanding into emerging research areas such as satellite video scene classification and object segmentation. However, there is no comprehensive review and summary in the intelligent processing of satellite video. This article presents a systematic review and quantitative analysis of the results published over the last decade, intending to further promote the development of various intelligent processing tasks for satellite video. It analyzes the current difficulties, challenges, and the methodological system for each task. In addition, it provides an in-depth analysis and summary of publicly available datasets and evaluation benchmarks for each task, as well as classic algorithm performance and application scenarios. Finally, this article summarizes the current research status and looks forward to the future development trend, hoping to inspire researchers in related fields and jointly promote the development of intelligent processing of satellite video.

Few-shot learning, which aims to learn the concept of novel category from extremely limited labeled samples, has received intense interests in remote sensing image scene classification. Most of the existing methods inherit the philosophy of prototype learning and tackle classification as the prototype-based metric matching problem. Despite the achievement that has been obtained so far, the problems of interclass metric misalignment and intraclass variations have become two main challenges that obstacle the performance. In this article, a novel transductive learning framework with conditional metric embedding is proposed to remedy these problems. First, a conditional metric embedding mechanism is introduced to perform anisotropic embedding for each pair of the support category and query instance. This design provides the model with flexible scalability to accommodate the metric biases across classes. Second, a transductive prototype learning strategy is presented to enhance the robustness of the prototype against intraclass variations. The unlabeled query instances are transformed into pseudo instances equipped with credibility coefficients and then leveraged to calibrate the prototype estimation bias in low-data regimes. Third, a long-term consistency regularization is designed, which continuously memorizes the historical prototypes to generate additional supervision of interclass separation in the global label space. Benefiting from this design, the discriminability of the prototypes obtains obvious improvement. Finally, extensive experiments are conducted on three public benchmark remote sensing datasets. The experimental results demonstrate the validity and the superiority of the proposed method in low-shot conditions.

The rapid development of remote sensing (RS) technology has produced massive images, which makes it difficult to obtain interpretation results by manual screening. Therefore, researchers began to develop automatic retrieval method of RS images. In recent years, cross-modal RS image retrieval based on query text has attracted many researchers because of its flexible and has become a new research trend. However, the primary problem faced is that the information of query text and RS image is not aligned. For example, RS images often have the attributes of multiscale and multiobjective, and the amount of information is rich, while the query text contains only a few words, and the information is scarce. Recently, graph neural network (GNN) has shown its potential in many tasks with its powerful feature representation ability. Therefore, based on GNN, this article proposes a new cross-modal RS feature matching network, which can avoid the degradation of retrieval performance caused by information misalignment by learning the feature interaction in query text and RS image, respectively, and modeling the feature association between the two modes. Specifically, to fuse the within-modal features, the text and RS image graph modules are designed based on GNN. In addition, in order to effectively match the query text and RS image, combined with the multihead attention mechanism, an image-text association module is constructed to focus on the parts related to RS image in the text. The experiments on two public standard datasets verify the competitive performance of the proposed model.

Cross-modal remote sensing (RS) image retrieval aims to retrieve RS images using other modalities (e.g., text) and vice versa. The relationship between objects in the RS image is complex, i.e., the distribution of multiple types of objects is uneven, which makes the matching with query text inaccurate, and then restricts the performance of remote sensing image retrieval. Previous methods generally focus on the feature matching between RS image and text and rarely model the relationships between features of RS image. Hypergraph (hyperedge connecting multiple vertices) is an extended structure of a regular graph and has attracted extensive attention for its superiority in representing high-order relationships. Inspired by the advantages of the hypergraph, in this work, a hypergraph-enhanced textual-visual matching network (HyperMatch) is proposed to circumvent the inaccurate matching between the RS image and query text. Specifically, a multiscale RS image hypergraph network is designed to model the complex relationships between features of the RS image for forming the valuable and redundant features into different hyperedges. In addition, a hypergraph construction and update method for an RS image is designed. For constructing a hypergraph, the features of an RS image running as vertices and cosine similarity is the metric to measure the correlation between them. Vertex and hyperedge attention mechanisms are introduced for the dynamic update of a hypergraph to realize the alternating update of vertices and hyperedges. Quantitative and qualitative experiments on the RSICD and RSITMD datasets verify the effectiveness of the proposed method in cross-modal remote sensing image retrieval.

LaMnO3 is one of the most common Mn-based perovskites which has been studied as a low-temperature catalyst for NOx reduction. However, the Hg0 oxidation performance of LaMnO3 is not ideal. An acid treatment method was adopted to regulate the LaMnO3 surface to expose more active sites. The obtained MnO2/ LaMnO3 sample showed enhanced SCR and Hg0 oxidation performances. The NO conversion rate and Hg0 removal efficiency of MnO2/ LaMnO3 could be over 85% in the temperature range of 150−250 °C. The results suggested that the catalyst might be used to remove NO and Hg0 simultaneously at low temperature in coal-fired flue gas.

Abstract The transformer architecture has been the dominant framework for today's image captioning tasks because of its superior performance. However, existing methods based on transformer often lack the integrated use of multi‐level semantic information and are weak in maintaining the relevance of captions to the image. In this paper, a semantic‐meshed and content‐guided transformer network is introduced for image captioning to solve these problems. The semantic‐meshed mechanism allows the model to generate words by selecting semantic information of multiple interaction levels adaptively through attention‐based reconstruction. And the content‐guided module guides the words generation by using attribute features that represent the image content, which aims to keep the generated caption consistent with the main content of the image. Experiments on dataset on the MSCOCO captioning dataset are conducted to validate the authors’ model and achieve superior results compared to other state‐of‐the‐art method approaches.

The sample scarcity issue is still challenged for SAR images interpretation. The number of geospatial targets related images is constrained of the SAR images interpretation ability of data acquisition, sample labeling, and the lack of target coverage. Our SAR-ATR method is demonstrated based on scattering information and meta-learning. First, the discrete distribution of the spatial structure of different types of aircraft is quite different in SAR images. An associated scattering classifier is designed to guide the network to learn more discriminative intra-class and inter-class feature descriptions. Our proposed classifier facilitates the modeling of discrete degree of the aircraft target quantitatively and balance the weights of sample pairs dynamically through the differentiated analysis of different target discrete distributions. In addition, an adaptive feature refinement module is designed to optimize the network cohesion for the key parts of the aircraft and reduce the interference of background noise. The proposed method integrates the target scattering distribution properties to the network learning process. On 5-way 1-shot emerging categorized recognition task involved only few samples, our experimental results demonstrate that the recognition accuracy of this method is 59.90%, which is 3.85% higher than the benchmark. After reducing the amount of training data by half, the proposed method is still competitive on the new category of few-shot recognition tasks.

Porcine deltacoronavirus (PDCoV) is an emerging virus that poses a significant threat to the global swine industry. Its membrane (M) protein is crucial for virion assembly and virus–host interactions. We selected the hydrophilic region of M protein for prokaryotic expression, purification, and recombinant protein production. Utilizing hybridoma technology, we prepared the monoclonal antibody (mAb) 24-A6 against M protein. The mAb 24-A6 was shown to be suitable for use in immunofluorescence assays, western blotting, and immunoprecipitation, with specificity for PDCoV and no cross-reactivity with other five porcine viruses. The M protein was observed to be expressed as early as 3 h after PDCoV infection, increasing its expression over the duration of infection. Notably, the antigenic epitope of the M protein identified as 103SPESRL108 recognized by mAb 24-A6 was found within a conserved structural domain (SWWSFNPETNNL) of the coronavirus M protein, indicating a crucial overlap between a functionally important viral assembly region and a region recognized by the immune system. Our findings provide valuable insights into mAb 24-A6 targeting the antigenic epitope of M protein and may contribute to the development of diagnostic tools for PDCoV infection and fundamental research into the function of PDCoV M protein.

Abstract The current image captioning methods usually integrate an object detection network to obtain image features at the level of objects and other salient regions. However, the detection network needs to be independently pre‐trained on additional data. Thus, mainly due to the demand for extra training data and computing resources, the detection network's utilization will impose higher training costs on the overall captioning model. In this work, the authors propose a local–global attention model based on two‐step clustering features for image captioning. The two‐step clustering features can be obtained at a relatively low cost and have the presentation ability in objects or other salient image regions. To make the model perceive the image better, the authors introduce a novel local–global attention mechanism. The model will analyse the clustering features from local perspectives to global ones at each time step, making the model better understand the image contents. The authors evaluate the proposed method on the MSCOCO test server, achieving BLEU‐4/METEOR/ROUGE‐L scores of 36.8, 27.4, and 57.2, respectively. With the benefit of reducing training costs, the authors' model also achieves closing results compared with the models using detection features.

The relationship between metal vapor in arc and droplet transfer mode in GMAW was reviewed. The results show that when welding with high current and corresponding welding parameters and different shielding gases, the conductivity of the vapor in arc plasma center is different, resulting in different positions of current path, and finally leading to different (streaming, projected, and droplet) droplet transfer modes. The metal vapor reduces the electrical conductivity at the center of arc plasma, expands the current path in the arc, and the direction of the dominant force acting on droplet facilitates droplet transfer, which fully meets the formation conditions of droplet spray transfer. The effects of shielding gas type, arc temperature, and welding wire composition on metal vapor mainly depend on their effects on the electrical conductivity change of arc plasma. It is possible to regulate the droplet transfer phenomenon by controlling the conductivity of metal vapor to realize the spray transfer of pure CO2 shielding gas. Further research is needed on the key technologies.

The aim of this paper is to examine how managerial ability affects corporate tax aggressiveness. The study follows the work of Demerjian, Lev, and McVay (2012) and quantifies managerial ability by calculating how efficiently managers generate revenues from given economic resources using the data envelopment analysis (DEA) approach. The study uses a wide range of measures of tax aggressiveness. Firm fixed-effects regressions and a difference-in-differences approach using information regarding CEO turnover to control for endogeneity are used. The study finds a negative relationship between managerial ability and corporate tax aggressiveness. Further tests show that this negative relationship is more pronounced for firms with higher investment opportunities or firms with more reputational concerns. Given the significant costs associated with tax aggressiveness and the negative effect it can have on managerial reputation if discovered, the results suggest that more able managers invest less effort in aggressive tax avoidance activities. This study furthers the understanding of how managerial personal traits affect corporate decision-making.

Dredged sandy clays are widely utilized as fill material in geotechnical projects, and their compressibility is crucial to the engineering design. It is well recognized that the compression behavior of dredged sandy clays is significantly affected by the initial water content and sand fraction. Besides, these two physical properties are usually spatially dependent due to the process of hydraulic filling. In this study, an effective compression model is formulated based on the equivalent void ratio concept. The compression curve of pure clays is first established as a reference, which depends on the initial void ratio and liquid limit. Then, the effect of sand fraction has been incorporated by directly implementing the equivalent void ratio into the reference curve. A structural variable is introduced to describe the evolving intergranular structure with coarse content and stress level. Only one additional structural parameter is required, and it can be readily calibrated by one traditional compression test. Validation shows that the model is effective in reproducing the compression behavior of dredged sandy clays with a wide range of initial water content and sand fraction. The proposed equation can provide a reasonable reference for establishing the hardening law and the corresponding general constitutive model of dredged sandy clays.

The bans on bisphenol A (BPA) have facilitated the widespread use of BPA alternatives and shifted environmental contamination profiles of bisphenols (BPs). However, the continued reports of toxicities of emerging BPA alternatives have raised questions about whether the shifting profiles are contributed to mitigate BPs-mediated endocrine-disruption effects (EDEs). Cetaceans are commonly used as the ideal sentinel species for monitoring marine pollutants of concern and determining potential health effects, but far less is known about BP loads and BPs-mediated EDEs in cetaceans. Here we measured the hepatic concentrations of six BPs in eight stranded cetacean species (n = 41) in the South China Sea, between 2007 and 2020. The large-bodied whales generally showed higher ∑BPs concentrations than the small-bodied dolphins. In Indo-Pacific finless porpoises (Neophocaena phocaenoides) (n = 33), BPA concentrations first increased (2007–2014) and then decreased (2014–2020), while ∑BPAlternatives concentrations increased from 2007 to 2020. It appears that the alternatives gradually replaced BPA, probably due to the BPA-related bans in China. In order to examine the hormone disruption of BPA and its alternatives in finless porpoises, five blubber hormones (cortisol, progesterone, testosterone, triiodothyronine and tetraiodothyronine), which are proven to be validated endocrine biomarkers, were measured in 21 samples. Tetraiodothyronine, testosterone, and cortisol were significantly and positively correlated with BPA and its alternatives, suggesting that the interference of endocrine hormone homeostasis may continue to occur despite the changes of BP profiles in finless porpoises. This is the first investigation of the relationship between hormone and BP concentrations in cetaceans and represents a substantial advance in understanding BPs-mediated endocrine effects on cetaceans.

Cross-modal remote sensing retrieval (RSCR) has an increasing importance due to the ability to quickly and flexibly retrieve valuable data from enormous remote sensing (RS) images. However, traditional RSCR methods tend to focus on the retrieval between two modalities, when the number of modalities increases, the contradiction between the increasing semantic gap and the small amount of paired data causes the model fail to learn a superior modal representation. In this paper, inspired by the visual-based modal center in RS, we construct a multi-source cross-modal retrieval network (MCRN) that manages to unify RS retrieval tasks under multiple retrieval sources. To solve the data heterogeneity caused by multiple data sources, we propose a shared pattern transfer module (SPTM) based on pattern memory and combine the theory of generative adversarial to achieve the semantic representation unbound from modality. Simultaneously, to cope with the lack of annotation data in the RS scenario, multiple unimodal self-supervised frameworks are unified to obtain robust pre-training parameters for the designed MCRN by combining domain alignment and contrastive learning. Finally, we come up with the multi-source triplet loss, the unimodal contrast loss, and the semantic consistency loss, which efficaciously make MCRN achieve competitive results through multitask learning for semantic alignment. We construct multimodal datasets M-RSICD and M-RSITMD, conduct extensive experiments and provide a complete benchmark to facilitate the development of RS multi-source cross-modal retrieval. The code of the MCRN method and the proposed dataset have been open to access at [Link].

Over the recent years, high-resolution Synthetic-Aperture Radar (SAR) images have been widely applied for intelligent interpretation of urban mapping, change detection, etc. Different from optical images, the acquisition approach and object geometry of SAR images have limited the interpretation performances of the existing deep-learning methods. This paper proposes a novel building footprint generation method for high-resolution SAR images. This method is based on supervised contrastive learning regularization, which aims to increase the similarities between intra-class pixels and diversities of interclass pixels. This increase will make the deep learning models focus on distinguishing building and nonbuilding pixels in latent space, and improve the classification accuracy. Based on public SpaceNet6 data, the proposed method can improve the segmentation performance by 1% compared to the other state-of-the-art methods. This improvement validates the effectiveness of the proposed method on real data. This method can be used for building segmentation in urban areas with complex scene background. Moreover, the proposed method can be extended for other types of land-cover segmentation using SAR images.

With the development of deep learning, great progress has been made in object detection of remote sensing (RS) imagery. However, the object detector is hard to generalize well from one labeled dataset (source domain) to another unlabeled dataset (target domain) due to the discrepancy of data distribution (domain shift). Currently, adversarial-based domain adaptation methods align the semantic features of source and target domain features to alleviate the domain shift. But they fail to avoid the alignment of noisy background features and neglect the instance-level features, which are inappropriate for detection models that focus on instance location and classification. To mitigate domain shift existing in object detection, we propose a reconstructed feature alignment network (RFA-Net) for unsupervised cross-domain object detection in RS imagery. The RFA-Net includes one sequential data augmentation module deployed on data level for providing solid gains on unlabeled data, one sparse feature reconstruction module deployed on feature level to intensify instance feature for feature alignment, and one pseudo-label generation module deployed on label level for the supervision of the unlabeled target domain. Extensive experiments illustrate that our proposed RFA-Net is effective to alleviate the domain shift problem in domain adaptation object detection of RS imagery.

Airborne laser scanning point cloud semantic labeling, which aims to identify the category of each point, plays a significant role in many applications, such as forest observing, powerline extraction, etc. Under the guidance of deep learning technology, the interpretation thought of point clouds has also greatly changed. However, owing to the irregular and unordered natures of point clouds, it is relatively difficult for classification model to distinguish some objects with similar geometry by single-modal data only. Fortunately, additional gain information, e.g., color spectrum which can be complementary to geometric information, is able to effectively promote the classification effect. Therefore, the design of fusion strategy is a critical part in model construction. In this article, aiming to capture more abstract semantic information for color spectrum data, we elaborate a color spectrum fusion (CSF) module. It can be flexibly integrated into a classification pipeline with just negligible parameters. Then, we expand data fusion thoughts for point clouds and color spectrum and investigate three possible fusion strategies. Accordingly, we develop three architectures to construct CSF-Nets. Ultimately, by taking a weighted cross entropy loss, we can train our CSF-Nets in an end-to-end manner. Experiments on two extensively used datasets: Vaihingen 3D and LASDU show that the presented three fusion approaches all can improve the performance, while the earlier fusion strategy performs the best. Besides, compared with other well-performed methods, CSF-Net is still able to achieve satisfactory performance on overall accuracy and m$F_{1}$-score indicator. This also validates the effectiveness of our multimodal fusion network.

Changes in structure of oral solid dosage forms (OSDF) elementally determine the drug release and its therapeutic effects. In this research, synchrotron radiation X-ray micro-computed tomography was utilized to visualize the 3D structure of enteric coated pellets recovered from the gastrointestinal tract of rats. The structures of pellets in solid state and in vitro compendium media were measured. Pellets in vivo underwent morphological and structural changes which differed significantly from those in vitro compendium media. Thus, optimizations of the dissolution media were performed to mimic the appropriate in vivo conditions by introducing pepsin and glass microspheres in media. The sphericity, pellet volume, pore volume and porosity of the in vivo esomeprazole magnesium pellets in stomach for 2 h were recorded 0.47, 1.55 × 108 μm3, 0.44 × 108 μm3 and 27.6%, respectively. After adding pepsin and glass microspheres, the above parameters in vitro reached to 0.44, 1.64 × 108 μm3, 0.38 × 108 μm3 and 23.0%, respectively. Omeprazole magnesium pellets behaved similarly. The structural features of pellets between in vitro media and in vivo condition were bridged successfully in terms of 3D structures to ensure better design, characterization and quality control of advanced OSDF.

Ship detection and classification in synthetic aperture radar (SAR) images play a vital role for wide applications. Due to the unique SAR imaging mechanism, ship detection and classification tasks have faced numerous challenges, such as land interference, image defocus, and noise. Many detectors and classifiers have been presented to handle these problems. However, the general deep learning-based detectors and classifiers lack the combination of SAR characteristics, which leads to poor performance. Compared with optical images, SAR images lack the texture information of ships, which brings great difficulties to the recognition task. To address the above issues, a novel deep learning-based ship detection and classification network combined with scattering characteristics is proposed in this article. First, to accurately locate ships in large-scale SAR images, this article designs a strong scattering point aware network (SPAN) by capturing the strong scattering points that existed in the ship area. SPAN recognizes the ship category according to their distribution characteristics. Second, to compensate for the feature loss caused by the down-sampling operation, this article designs a more suitable resolution recovery module to replace the bilinear interpolation method. Third, a region of interest automatic generation module is proposed to fully utilize the axis-align feature of oriented proposal boxes and the sufficient information of horizontal proposal boxes. Furthermore, the classification encoder module extracts the distribution feature of scattering points to classify SAR ships. Finally, the comprehensive experiments in the large-scale dataset for ship detection and classification in SAR images (LDSD) demonstrate the superior performance of the proposed method.

Much of the recent work in remote sensing image captioning is influenced by natural image captioning. These methods tend to fix the defects of the model architecture to improve the previous work, but pay little attention to the differences between remote sensing images and natural images. By considering these differences, we propose a multiscale multiinteraction remote sensing image captioning model. As in Fig. 1(a), the targets in remote sensing images have a wide range of scales; while the natural images are generally taken close-up, resulting in a similar scale for the foreground targets. Due to the difference in shooting methods, the model pretrained on close-up natural images cannot capture multiscale remote sensing targets well. To alleviate this problem, we propose a two-stage multiscale structure for feature representation, where we first finetune the CNN backbone on remote sensing images for domain adaption, then we collect features from different stages as the multiscale feature representation. Moreover, due to the shooting distance, the height information of the target in the remote sensing image is greatly weakened, thus some objects like low plants and grasses become difficult to identify, as in Fig. 1(b). Thus, we further propose a multiinteraction feature representation module, where information flow of the same and different layers could effectively interact. By calculating the similarity score among features, we fuse features with high similarity, and increase the distance between features of different categories, thereby enhancing the distinguishability. Results on RSICD, Sydney-Captions, and UCM-Captions show a clear improvement over the compared methods.

Polarimetric synthetic aperture radar (PolSAR) terrain segmentation is a fundamental research topic in PolSAR image interpretation. Recently, many studies have been investigated to handle this task. However, the existing data for PolSAR terrain segmentation have relatively limited scale and their scene complexity is relatively simple. These issues greatly restrict the development of algorithms. Therefore, there is a strong requirement for establishing a large-scale data set for terrain segmentation in complex-scene PolSAR images. In this paper, we present a benchmark data set containing a PolSAR amplitude image with a 9082×9805-pixel region and 2000 image patches with a size of 512×512 for PolSAR terrain segmentation, which is called AIR-PolSAR-Seg. We collect the PolSAR image with a resolution of 8 m from the GaoFen-3 satellite, and it is equipped with pixel-wise annotation which covers six categories. Compared with the previous data resources, AIR-PolSAR-Seg preserves some specific properties. First, AIR-PolSAR-Seg owns a large-size PolSAR image and provides a large quantity of image patches. It offers the research community a complete data resource with adequate training examples and reliable validation results. Second, AIR-PolSAR-Seg is established upon a PolSAR image with high scene complexity. This characteristic motivates robust and advanced segmentation approaches to facilitate complex-scene PolSAR image analysis. Based on AIR-PolSAR-Seg, three tasks are introduced: multi-category segmentation, water body segmentation, and building segmentation. Moreover, a performance analysis of traditional approaches and deep learning-based approaches are conducted, which can be regarded as baselines and provide references for future research.

Chronic kidney disease (CKD) is often accompanied with imbalanced gut microbiota and impaired intestinal barrier. Hence, efforts to ameliorate renal dysfunction by manipulating gut microbial ecosystem are underway. Yishen Qingli Heluo granule (YQHG) is a representative traditional Chinese medicine (TCM) prescription for clinical treatment of CKD. However, its underlying mechanism has not been well elucidated. This study aimed to explore effects of YQHG on renal dysfunction in 5/6 nephrectomized rats by targeting gut microbiota and intestinal barrier. Here, we found that YQHG provided significant renal protection in 5/6 nephrectomized rats by reducing renal fibrosis and inflammation, reestablishing bacterial communities, and improving intestinal barrier. Our analysis showed that YQHG altered the bacterial community of 5/6 nephrectomized rats. In particular, the prescription significantly increased the relative abundance of SCFA-producing bacteria (i.e., Lactobacillaceae, Lactobacillus and Lactobacillus_gasseri), which was contributed to the improved SCFA concentration (i.e., total SCFA, acetic acid, butyric acid) and intestinal barrier (i.e., the improved permeability and microbial translocation). More critically, microbiota-transfer study showed that the protective effect of YQHG was partly attributed to the mediation of the gut microbiota, especially the SCFA-producing bacteria. Our current findings propose a microbiota-targeted intervention and indicate that YQHG may become a novel promising treatment for CKD.

Abstract Background Gestational diabetes mellitus (GDM) is a metabolic disease that occurs in pregnant women and increases the risk for the development of diabetes. The relationship between GDM and meconium microbiota and metabolome remains incompletely understood. Methods Four hundred eighteen mothers (147 women with GDM and 271 normal pregnant women) and their neonates from the GDM Mother and Child Study were included in this study. Meconium microbiota were profiled by 16S rRNA gene sequencing. Meconium and maternal serum metabolome were examined by UPLC-QE. Results Microbial communities in meconium were significantly altered in neonates from the GDM mothers. A reduction in alpha diversity was observed in neonates of GDM mothers. At the phylum level, the abundance of Firmicutes and Proteobacteria changed significantly in neonates of GDM mothers. Metabolomic analysis of meconium showed that metabolic pathways including taurine and hypotaurine metabolism, pyrimidine metabolism, beta-alanine metabolism, and bile acid biosynthesis were altered in GDM subjects. Several changed metabolites varying by the similar trend across the maternal serum and neonatal meconium were observed. Conclusion Altogether, these findings suggest that GDM could alter the serum metabolome and is associated with the neonatal meconium microbiota and metabolome, highlighting the importance of maternal factors on early-life metabolism.

Abstract Genetic differentiations and phylogeographical patterns of small organisms may be shaped by spatial isolation, environmental gradients, and gene flow. However, knowledge about genetic differentiation of rotifers at the intercontinental scale is still limited. Polyarthra dolichoptera and P. vulgaris are cosmopolitan rotifers that are tolerant to environmental changes, offering an excellent model to address the research gap. Here, we investigated the populations in Southeastern China and eastern North America and evaluated the phylogeographical patterns from their geographical range sizes, geographic–genetic distance relationships and their responses to spatial‐environmental factors. Using the mitochondrial cytochrome c oxidase subunit I gene as the DNA marker, we analyzed a total of 170 individuals. Our results showed that some putative cryptic species, also known as entities were widely distributed, but most of them were limited to single areas. The divergence of P. dolichoptera and P. vulgaris indicated that gene flow between continents was limited while that within each continent was stronger. Oceanographic barriers do affect the phylogeographic pattern of rotifers in continental waters and serve to maintain genetic diversity in nature. The genetic distance of P. dolichoptera and P. vulgaris populations showed significant positive correlation with geographic distance. This might be due to the combined effects of habitat heterogeneity, long‐distance colonization, and oceanographic barriers. Furthermore, at the intercontinental scale, spatial distance had a stronger influence than environmental variables on the genetic differentiations of both populations. Wind‐ and animal‐mediated transport and even historical events of continental plate tectonics are potential factors for phylogeography of cosmopolitan rotifers.

Simultaneously visualizing Amyloid-β (Aβ) plaque with its surrounding brain structures at the subcellular level in the intact brain is essential for understanding the complex pathology of Alzheimer's disease, but is still rarely achieved due to the technical limitations. Combining the micro-optical sectioning tomography (MOST) system, whole-brain Nissl staining, and customized image processing workflow, we generated a whole-brain panorama of Alzheimer's disease mice without specific labeling. The workflow employed the steps that include virtual channel splitting, feature enhancement, iso-surface rendering, direct volume rendering, and feature fusion to extract and reconstruct the different signals with distinct gray values and morphologies. Taking advantage of this workflow, we found that the denser-distribution areas of Aβ plaques appeared with relatively more somata and smaller vessels, but show a dissimilar distributing pattern with nerve tracts. In addition, the entorhinal cortex and adjacent subiculum regions present the highest density and biggest diameter of plaques. The neuronal processes in the vicinity of these Aβ plaques showed significant structural alternation such as bending or abrupt branch ending. The capillaries inside or adjacent to the plaques were observed with abundant distorted micro-vessels and abrupt ending. Depicting Aβ plaques, somata, nerve processes and tracts, and blood vessels simultaneously, this panorama enables us for the first time, to analyze how the Aβ plaques interact with capillaries, somata, and processes at a submicron resolution of 3D whole-brain scale, which reveals potential pathological effects of Aβ plaques from a new cross-scale view. Our approach opens a door to routine systematic studies of complex interactions among brain components in mouse models of Alzheimer's disease.

Galectin-9 regulates several cellular processes including TIM-3-mediated T cell death. Here the authors show that co-expressed PD-1 protects TIM-3+ T cells from galectin-9-induced cell death and that anti-galectin-9 in combination with GITR agonism promotes an anti-tumor immune response.

The detection of Thermal Power Plants (TPPs) is a meaningful task for remote sensing image interpretation. It is challenging due to the variations in appearance and complex structures. In this article, we propose a novel end-to-end detection framework for TPPs based on deep convolutional neural networks. Specifically, a large-scale TPPs Dataset for Detection (AIR-TPPDD) in remote sensing images is presented. AIR-TPPDD is collected from the Google Earth worldwide, and provides detailed annotations including names and locations. To the best of our knowledge, this is the first publicly available dataset for TPP detection. Then, based on Faster R-CNN, a saliency enhanced module is proposed to strengthen the ability in representing complex structure, as well as alleviate distractions in the background. In addition, we design a multi-scale feature module to adapt to the large size range of TPPs. Experiments show that the proposed method outperforms the state-of-the-art methods and achieves 76.7% mAP on the challenging AIR-TPPDD.

The rapid development of deep learning brings effective solutions for remote sensing image interpretation. Training deep neural network models usually require a large number of manually labeled samples. However, there is a limitation to obtain sufficient labeled samples in remote sensing field to satisfy the data requirement. Therefore, it is of great significance to conduct the research on few-shot learning for remote sensing image interpretation. First, this article provides a bibliometric analysis of the existing works for remote sensing interpretation related to few-shot learning. Second, two categories of few-shot learning methods, i.e., the data-augmentation-based and the prior-knowledge-based, are introduced for the interpretation of remote sensing images. Then, three typical remote sensing interpretation applications are listed, including scene classification, semantic segmentation, and object detection, together with the corresponding public datasets and the evaluation criteria. Finally, the research status is summarized, and some possible research directions are provided. This article gives a reference for scholars working on few-shot learning research in the remote sensing field.

The building extraction from synthetic aperture radar (SAR) images has always been a challenging research topic. Recently, the deep convolution neural network brings excellent improvements in SAR segmentation. The fully convolutional network and other variants are widely transferred to the SAR studies because of their high precision in optical images. They are still limited by their processing in terms of the geometric distortion of buildings, the variability of building structures, and scattering interference between adjacent targets in the SAR images. In this article, a unified framework called selective spatial pyramid dilated (SSPD) network is proposed for the fine building segmentation in SAR images. First, we propose a novel encoder–decoder structure for the fine building feature reconstruction. The enhanced encoder and the dual-stage decoder, composed of the CBM and the SSPD module, extract and recover the crucial multiscale information better. Second, we design the multilayer SSPD module based on the selective spatial attention. The multiscale building information with different attention on multiple branches is combined, optimized, and adaptively selected for adaptive filtering and extracting features of complex multiscale building targets in SAR images. Third, according to the building features and SAR imaging mechanism, a new loss function called L-shape weighting loss (LWloss) is proposed to heighten the attention on the L-shape footprint characteristics of the buildings and reduce the missing detection of line buildings. Besides, LWloss can also alleviate the class imbalance problem in the optimization stage. Finally, the experiments on a large-scene SAR image dataset demonstrate the effectiveness of the proposed method and verify its superiority over other approaches, such as the region-based Markov random field, U-net, and DeepLabv3+.

Point cloud classification is regarded as a critical task in remote sensing data interpretation, which is widely used in many fields. Recently, many proposed methods tend to develop an end-to-end network to directly operate on the raw point cloud, which has shown great power. However, most of these methods abstract local features by equally considering the neighboring points. The features learned may neglect to distinguish contributions of different points especially the edge points and outliers, leading to a coarse classification result especially for boundaries. Moreover, the extracted features are high redundant and intercorrelated with similar categories, posing difficulty in identifying classes sharing similar characteristics especially in complex scenes. Therefore, we propose an adaptive context balancing and feature filtering network (CBF-Net) to tackle the aforementioned problems. First, we introduce a balanced context encoding module to balance semantically the features of neighboring points, which can help the model learn more from the edge points and, therefore, contribute to a finer classification. Then, considering that the interference for similar classes probably causes confusion among them, a filtered feature aggregating module is proposed to filter the extracted features by mapping them into a cleaner subspace with a lower rank. We have conducted thorough experiments on the International Society for Photogrammetry and Remote Sensing 3-D labeling dataset. Experimental results show that our CBF-Net can obtain high accuracy and achieve state-of-the-art level in the categories of Powerline, Car, and Facade. In addition, we also conduct experiments on the RueMonge2014 dataset, which further reveals the strong ability of our model.

In this article, we introduce the 2020 Gaofen Challenge and relevant scientific outcomes. The 2020 Gaofen Challenge is an international competition, which is organized by the China High-Resolution Earth Observation Conference Committee and the Aerospace Information Research Institute, Chinese Academy of Sciences and technically cosponsored by the IEEE Geoscience and Remote Sensing Society and the International Society for Photogrammetry and Remote Sensing. It aims at promoting the academic development of automated high-resolution earth observation image interpretation. Six independent tracks have been organized in this challenge, which cover the challenging problems in the field of object detection and semantic segmentation. With the development of convolutional neural networks, deep-learning-based methods have achieved good performance on image interpretation. In this article, we report the details and the best-performing methods presented so far in the scope of this challenge.

Semantic segmentation in aerial images has become an indispensable part in remote sensing image understanding for its extensive application prospects. It is crucial to jointly reason the 2-D appearance along with 3-D information and acquire discriminative global context to achieve better segmentation. However, previous approaches require accurate elevation data (e.g., nDSM and Digital Surface Model (DSM)) as additional inputs to segment semantics, which sorely limits their applications. On the other hand, due to the various forms of objects in complex scenes, the global context is generally dominated by features of salient patterns (e.g., large objects) and tends to smooth inconspicuous patterns (e.g., small stuff and boundaries). In this article, a novel joint framework named height-embedding context reassembly network (HECR-Net) is proposed. First, considering the fact that the corresponding elevation data is insufficient while we still want to exploit the serviceable height information, to alleviate the above data constraint, our method simultaneously predicts semantic labels and height maps from single aerial images by distilling height-aware embeddings implicitly. Second, we introduce a novel context-aware reorganization module to generate a discriminative feature with global context appropriately assigned to each local position. It benefits from both the global context aggregation module for ambiguity eliminating and local feature redistribution module for detailed refinement. Third, we make full use of the learning height-aware embeddings to promote the performance of semantic segmentation via introducing a modality-affinitive propagation block. Finally, without bells and whistles, the segmentation results on ISPRS Vaihingen and Potsdam data set illustrate that the proposed HECR-Net achieves state-of-the-art performance.