scholarly journals UATNet: U-Shape Attention-Based Transformer Net for Meteorological Satellite Cloud Recognition

2021 ◽  
Vol 14 (1) ◽  
pp. 104 ◽  
Author(s):  
Zhanjie Wang ◽  
Jianghua Zhao ◽  
Ran Zhang ◽  
Zheng Li ◽  
Qinghui Lin ◽  
...  
Keyword(s):  
Weather Forecasting ◽  
Long Distance ◽  
Coverage Area ◽  
Structural Shape ◽  
Multi Scale ◽  
Basic Task ◽  
Satellite Cloud Image ◽  
Long Time ◽  
Channel Information ◽  
Meteorological Satellite

Cloud recognition is a basic task in ground meteorological observation. It is of great significance to accurately identify cloud types from long-time-series satellite cloud images for improving the reliability and accuracy of weather forecasting. However, different from ground-based cloud images with a small observation range and easy operation, satellite cloud images have a wider cloud coverage area and contain more surface features. Hence, it is difficult to effectively extract the structural shape, area size, contour shape, hue, shadow and texture of clouds through traditional deep learning methods. In order to analyze the regional cloud type characteristics effectively, we construct a China region meteorological satellite cloud image dataset named CRMSCD, which consists of nine cloud types and the clear sky (cloudless). In this paper, we propose a novel neural network model, UATNet, which can realize the pixel-level classification of meteorological satellite cloud images. Our model efficiently integrates the spatial and multi-channel information of clouds. Specifically, several transformer blocks with modified self-attention computation (swin transformer blocks) and patch merging operations are used to build a hierarchical transformer, and spatial displacement is introduced to construct long-distance cross-window connections. In addition, we introduce a Channel Cross fusion with Transformer (CCT) to guide the multi-scale channel fusion, and design an Attention-based Squeeze and Excitation (ASE) to effectively connect the fused multi-scale channel information to the decoder features. The experimental results demonstrate that the proposed model achieved 82.33% PA, 67.79% MPA, 54.51% MIoU and 70.96% FWIoU on CRMSCD. Compared with the existing models, our method produces more precise segmentation performance, which demonstrates its superiority on meteorological satellite cloud recognition tasks.

scholarly journals Non-Local and Multi-Scale Mechanisms for Image Inpainting

Sensors ◽  
2021 ◽  
Vol 21 (9) ◽  
pp. 3281
Author(s):  
Xu He ◽  
Yong Yin
Keyword(s):  
Markov Random Fields ◽  
Receptive Fields ◽  
Image Inpainting ◽  
Long Distance ◽  
Visual Appearance ◽  
Fine Grained ◽  
Multi Scale ◽  
Non Local ◽  
Relationship Of

Recently, deep learning-based techniques have shown great power in image inpainting especially dealing with squared holes. However, they fail to generate plausible results inside the missing regions for irregular and large holes as there is a lack of understanding between missing regions and existing counterparts. To overcome this limitation, we combine two non-local mechanisms including a contextual attention module (CAM) and an implicit diversified Markov random fields (ID-MRF) loss with a multi-scale architecture which uses several dense fusion blocks (DFB) based on the dense combination of dilated convolution to guide the generative network to restore discontinuous and continuous large masked areas. To prevent color discrepancies and grid-like artifacts, we apply the ID-MRF loss to improve the visual appearance by comparing similarities of long-distance feature patches. To further capture the long-term relationship of different regions in large missing regions, we introduce the CAM. Although CAM has the ability to create plausible results via reconstructing refined features, it depends on initial predicted results. Hence, we employ the DFB to obtain larger and more effective receptive fields, which benefits to predict more precise and fine-grained information for CAM. Extensive experiments on two widely-used datasets demonstrate that our proposed framework significantly outperforms the state-of-the-art approaches both in quantity and quality.

scholarly journals Evolution of andrenine bees reveals a long and complex history of faunal interchanges between the Americas during the Mesozoic and Cenozoic

2021 ◽  
Author(s):  
Kelli S Ramos ◽  
Aline C Martins ◽  
Gabriel A R Melo
Keyword(s):  
North America ◽  
South America ◽  
Time Window ◽  
Distributional Pattern ◽  
Long Distance ◽  
Early Diversification ◽  
Palearctic Region ◽  
Long Time ◽  
The Times ◽  
Dated Phylogeny

Bees are presumed to have arisen in the early to mid-Cretaceous coincident with the fragmentation of the southern continents and concurrently with the early diversification of the flowering plants. Among the main groups of bees, Andreninae sensu lato comprise about 3000 species widely distributed with greatest and disjunct diversity in arid areas of North America, South America, and the Palearctic region. Here, we present the first comprehensive dated phylogeny and historical biogeographic analysis for andrenine bees, including representatives of all currently recognized tribes. Our analyses rely on a dataset of 106 taxa and 7952 aligned nucleotide positions from one mitochondrial and six nuclear loci. Andreninae is strongly supported as a monophyletic group and the recovered phylogeny corroborates the commonly recognized clades for the group. Thus, we propose a revised tribal classification that is congruent with our phylogenetic results. The time-calibrated phylogeny and ancestral range reconstructions of Andreninae reveal a fascinating evolutionary history with Gondwana patterns that are unlike those observed in other subfamilies of bees. Andreninae arose in South America during the Late Cretaceous around 90 Million years ago (Ma) and the origin of tribes occurred through a relatively long time-window from this age to the Miocene. The early evolution of the main lineages took place in South America until the beginning of Paleocene with North American fauna origin from it and Palearctic from North America as results of multiple lineage interchanges between these areas by long-distance dispersal or hopping through landmass chains. Overall, our analyses provide strong evidence of amphitropical distributional pattern currently observed in Andreninae in the American continent as result at least three periods of possible land connections between the two American landmasses, much prior to the Panama Isthmus closure. The andrenine lineages reached the Palearctic region through four dispersal events from North America during the Eocene, late Oligocene and early Miocene, most probably via the Thulean Bridge. The few lineages with Afrotropical distribution likely originated from a Palearctic ancestral in the Miocene around 10 Ma when these regions were contiguous, and the Sahara Desert was mostly vegetated making feasible the passage by several organisms. Incursions of andrenine bees to North America and then onto the Old World are chronological congruent with distinct periods when open-vegetation habitats were available for trans-continental dispersal and at the times when aridification and temperature decline offered favorable circumstances for bee diversification.

scholarly journals HYPERstream: a multi-scale framework for streamflow routing in large-scale hydrological model

2016 ◽  
Vol 20 (5) ◽  
pp. 2047-2061 ◽  
Author(s):  
Sebastiano Piccolroaz ◽  
Michele Di Lazzaro ◽  
Antonio Zarlenga ◽  
Bruno Majone ◽  
Alberto Bellin ◽  
...  
Keyword(s):  
Large Scale ◽  
Climate Models ◽  
Climate Model ◽  
Transfer Functions ◽  
Weather Forecasting ◽  
Computational Grid ◽  
River Network ◽  
Multi Scale ◽  
Continental Scale ◽  
Routing Scheme

Abstract. We present HYPERstream, an innovative streamflow routing scheme based on the width function instantaneous unit hydrograph (WFIUH) theory, which is specifically designed to facilitate coupling with weather forecasting and climate models. The proposed routing scheme preserves geomorphological dispersion of the river network when dealing with horizontal hydrological fluxes, irrespective of the computational grid size inherited from the overlaying climate model providing the meteorological forcing. This is achieved by simulating routing within the river network through suitable transfer functions obtained by applying the WFIUH theory to the desired level of detail. The underlying principle is similar to the block-effective dispersion employed in groundwater hydrology, with the transfer functions used to represent the effect on streamflow of morphological heterogeneity at scales smaller than the computational grid. Transfer functions are constructed for each grid cell with respect to the nodes of the network where streamflow is simulated, by taking advantage of the detailed morphological information contained in the digital elevation model (DEM) of the zone of interest. These characteristics make HYPERstream well suited for multi-scale applications, ranging from catchment up to continental scale, and to investigate extreme events (e.g., floods) that require an accurate description of routing through the river network. The routing scheme enjoys parsimony in the adopted parametrization and computational efficiency, leading to a dramatic reduction of the computational effort with respect to full-gridded models at comparable level of accuracy. HYPERstream is designed with a simple and flexible modular structure that allows for the selection of any rainfall-runoff model to be coupled with the routing scheme and the choice of different hillslope processes to be represented, and it makes the framework particularly suitable to massive parallelization, customization according to the specific user needs and preferences, and continuous development and improvements.

scholarly journals Identification of Metabolite and Lipid Profiles in a Segregating Peach Population Associated with Mealiness in Prunus persica (L.) Batsch

Metabolites ◽  
2020 ◽  
Vol 10 (4) ◽  
pp. 154 ◽  
Author(s):  
Victoria Lillo-Carmona ◽  
Alonso Espinoza ◽  
Karin Rothkegel ◽  
Miguel Rubilar ◽  
Ricardo Nilo-Poyanco ◽  
...  
Keyword(s):  
Cold Storage ◽  
Prunus Persica ◽  
Early Stage ◽  
Chilling Injury ◽  
Fruit Production ◽  
Lipid Profiles ◽  
Long Distance ◽  
Physiological Disorder ◽  
Export Markets ◽  
Long Time

The peach is the third most important temperate fruit crop considering fruit production and harvested area in the world. Exporting peaches represents a challenge due to the long-distance nature of export markets. This requires fruit to be placed in cold storage for a long time, which can induce a physiological disorder known as chilling injury (CI). The main symptom of CI is mealiness, which is perceived as non-juicy fruit by consumers. The purpose of this work was to identify and compare the metabolite and lipid profiles between two siblings from contrasting populations for juice content, at harvest and after 30 days at 0 °C. A total of 119 metabolites and 189 lipids were identified, which showed significant differences in abundance, mainly in amino acids, sugars and lipids. Metabolites displaying significant changes from the E1 to E3 stages corresponded to lipids such as phosphatidylglycerol (PG), monogalactosyldiacylglycerol (MGDG) and lysophosphatidylcholines (LPC), and sugars such as fructose 1 and 1-fructose-6 phosphate. These metabolites might be used as early stage biomarkers associated with mealiness at harvest and after cold storage.

scholarly journals Airborne Integrated Navigation System Based on SINS/GPS/TAN/EOAN

2020 ◽  
Vol 2020 ◽  
pp. 1-9
Author(s):  
Junjun Tang ◽  
Peijuan Li
Keyword(s):  
High Precision ◽  
Navigation System ◽  
Integrated Navigation ◽  
Long Distance ◽  
Integrated Navigation System ◽  
Long Time ◽  
Federated Kalman Filter ◽  
Entropy Weighted ◽  
Navigation Errors ◽  
Navigation Information

Considering the drawbacks that GPS signal is susceptible to obstacles and TAN becomes useless in some area when without any terrain data or with a featureless terrain field, to realize long-distance and high-precision navigation, a navigation system based on SINS/GPS/TAN/EOAN is presented. When GPS signal is available, GPS is used to correct errors of SINS; when GPS is unavailable, a terrain selection method based on the entropy weighted gray relational decision-making method is use to distinguish terrain into matchable areas and unmatchable areas; then, for the matchable areas, TAN is used to correct errors of SINS, for the unmatchable areas, EOAN is used to correct errors of SINS. The principles of SINS, GPS, TAN, and EOAN are analyzed, the mathematic models of SINS/GPS, SINS/TAN, and SINS/EOAN are constructed, and finally the federated Kalman filter is used to fuse navigation information. Simulation results show that the trajectory of SINS/GPS/TAN/EOAN is close to the ideal one in both matchable area or unmatchable area and whose navigation errors are obviously reduced, which is important for the realization of long-time and high-precision positioning.

scholarly journals Long Time Behaviour of an Exponential Integrator for a Vlasov-Poisson System with Strong Magnetic Field

2015 ◽  
Vol 18 (2) ◽  
pp. 263-296 ◽  
Author(s):  
Emmanuel Frénod ◽  
Sever A. Hirstoaga ◽  
Mathieu Lutz ◽  
Eric Sonnendrücker
Keyword(s):  
Time Behaviour ◽  
Poisson System ◽  
Particle In Cell ◽  
Multi Scale ◽  
Time Stepping ◽  
Long Time Behaviour ◽  
Dimensional Phase Space ◽  
Long Time ◽  
Exponential Integrator ◽  
Fast Oscillations

AbstractWith the aim of solving in a four dimensional phase space a multi-scale Vlasov-Poisson system, we propose in a Particle-In-Cell framework a robust time-stepping method that works uniformly when the small parameter vanishes. As an exponential integrator, the scheme is able to use large time steps with respect to the typical size of the solution’s fast oscillations. In addition, we show numerically that the method has accurate long time behaviour and that it is asymptotic preserving with respect to the limiting Guiding Center system.

Sign in / Sign up

Export Citation Format

Share Document