scholarly journals Deep Semantic-Preserving Reconstruction Hashing for Unsupervised Cross-Modal Retrieval

Entropy ◽  
2020 ◽  
Vol 22 (11) ◽  
pp. 1266
Author(s):  
Shuli Cheng ◽  
Liejun Wang ◽  
Anyu Du
Keyword(s):  
Large Scale ◽  
Semantic Information ◽  
Storage Capacity ◽  
Metric Learning ◽  
Model Parameters ◽  
Modal Interaction ◽  
Spatial Pooling ◽  
Visual Text ◽  
Global Covariance ◽  
Modal Semantic

Deep hashing is the mainstream algorithm for large-scale cross-modal retrieval due to its high retrieval speed and low storage capacity, but the problem of reconstruction of modal semantic information is still very challenging. In order to further solve the problem of unsupervised cross-modal retrieval semantic reconstruction, we propose a novel deep semantic-preserving reconstruction hashing (DSPRH). The algorithm combines spatial and channel semantic information, and mines modal semantic information based on adaptive self-encoding and joint semantic reconstruction loss. The main contributions are as follows: (1) We introduce a new spatial pooling network module based on tensor regular-polymorphic decomposition theory to generate rank-1 tensor to capture high-order context semantics, which can assist the backbone network to capture important contextual modal semantic information. (2) Based on optimization perspective, we use global covariance pooling to capture channel semantic information and accelerate network convergence. In feature reconstruction layer, we use two bottlenecks auto-encoding to achieve visual-text modal interaction. (3) In metric learning, we design a new loss function to optimize model parameters, which can preserve the correlation between image modalities and text modalities. The DSPRH algorithm is tested on MIRFlickr-25K and NUS-WIDE. The experimental results show that DSPRH has achieved better performance on retrieval tasks.

scholarly journals High-Rankness Regularized Semi-Supervised Deep Metric Learning for Remote Sensing Imagery

2020 ◽  
Vol 12 (16) ◽  
pp. 2603
Author(s):  
Jian Kang ◽  
Rubén Fernández-Beltrán ◽  
Zhen Ye ◽  
Xiaohua Tong ◽  
Pedram Ghamisi ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Large Scale ◽  
Semantic Information ◽  
Metric Spaces ◽  
Metric Learning ◽  
Training Data ◽  
Deep Metric Learning ◽  
Class Labels ◽  
Image Pairs ◽  
Decision Boundaries

Deep metric learning has recently received special attention in the field of remote sensing (RS) scene characterization, owing to its prominent capabilities for modeling distances among RS images based on their semantic information. Most of the existing deep metric learning methods exploit pairwise and triplet losses to learn the feature embeddings with the preservation of semantic-similarity, which requires the construction of image pairs and triplets based on the supervised information (e.g., class labels). However, generating such semantic annotations becomes a completely unaffordable task in large-scale RS archives, which may eventually constrain the availability of sufficient training data for this kind of models. To address this issue, we reformulate the deep metric learning scheme in a semi-supervised manner to effectively characterize RS scenes. Specifically, we aim at learning metric spaces by utilizing the supervised information from a small number of labeled RS images and exploring the potential decision boundaries for massive sets of unlabeled aerial scenes. In order to reach this goal, a joint loss function, composed of a normalized softmax loss with margin and a high-rankness regularization term, is proposed, as well as its corresponding optimization algorithm. The conducted experiments (including different state-of-the-art methods and two benchmark RS archives) validate the effectiveness of the proposed approach for RS image classification, clustering and retrieval tasks. The codes of this paper are publicly available.

scholarly journals Why ability point estimates can be pointless: a primer on using skill measures from large-scale assessments in secondary analyses

2021 ◽  
Vol 3 (1) ◽  
Author(s):  
Clemens M. Lechner ◽  
Nivedita Bhaktha ◽  
Katharina Groskurth ◽  
Matthias Bluemke
Keyword(s):  
Measurement Error ◽  
Statistical Models ◽  
Test Scores ◽  
Large Scale ◽  
Equation Modeling ◽  
Model Parameters ◽  
Advantages And Disadvantages ◽  
Point Estimates ◽  
Secondary Analyses ◽  
Large Scale Assessments

AbstractMeasures of cognitive or socio-emotional skills from large-scale assessments surveys (LSAS) are often based on advanced statistical models and scoring techniques unfamiliar to applied researchers. Consequently, applied researchers working with data from LSAS may be uncertain about the assumptions and computational details of these statistical models and scoring techniques and about how to best incorporate the resulting skill measures in secondary analyses. The present paper is intended as a primer for applied researchers. After a brief introduction to the key properties of skill assessments, we give an overview over the three principal methods with which secondary analysts can incorporate skill measures from LSAS in their analyses: (1) as test scores (i.e., point estimates of individual ability), (2) through structural equation modeling (SEM), and (3) in the form of plausible values (PVs). We discuss the advantages and disadvantages of each method based on three criteria: fallibility (i.e., control for measurement error and unbiasedness), usability (i.e., ease of use in secondary analyses), and immutability (i.e., consistency of test scores, PVs, or measurement model parameters across different analyses and analysts). We show that although none of the methods are optimal under all criteria, methods that result in a single point estimate of each respondent’s ability (i.e., all types of “test scores”) are rarely optimal for research purposes. Instead, approaches that avoid or correct for measurement error—especially PV methodology—stand out as the method of choice. We conclude with practical recommendations for secondary analysts and data-producing organizations.

scholarly journals Power-to-Green Methanol via CO2 Hydrogenation—A Concept Study Including Oxyfuel Fluidized Bed Combustion of Biomass

Energies ◽  
2021 ◽  
Vol 14 (15) ◽  
pp. 4638
Author(s):  
Simon Pratschner ◽  
Pavel Skopec ◽  
Jan Hrdlicka ◽  
Franz Winter
Keyword(s):  
Large Scale ◽  
Positive Impact ◽  
Renewable Energy Sources ◽  
Air Separation ◽  
Model Parameters ◽  
Processing Unit ◽  
Fluidized Bed Combustion ◽  
Separation Unit ◽  
Oxyfuel Combustion ◽  
Wind Park

A revolution of the global energy industry is without an alternative to solving the climate crisis. However, renewable energy sources typically show significant seasonal and daily fluctuations. This paper provides a system concept model of a decentralized power-to-green methanol plant consisting of a biomass heating plant with a thermal input of 20 MWth. (oxyfuel or air mode), a CO2 processing unit (DeOxo reactor or MEA absorption), an alkaline electrolyzer, a methanol synthesis unit, an air separation unit and a wind park. Applying oxyfuel combustion has the potential to directly utilize O2 generated by the electrolyzer, which was analyzed by varying critical model parameters. A major objective was to determine whether applying oxyfuel combustion has a positive impact on the plant’s power-to-liquid (PtL) efficiency rate. For cases utilizing more than 70% of CO2 generated by the combustion, the oxyfuel’s O2 demand is fully covered by the electrolyzer, making oxyfuel a viable option for large scale applications. Conventional air combustion is recommended for small wind parks and scenarios using surplus electricity. Maximum PtL efficiencies of ηPtL,Oxy = 51.91% and ηPtL,Air = 54.21% can be realized. Additionally, a case study for one year of operation has been conducted yielding an annual output of about 17,000 t/a methanol and 100 GWhth./a thermal energy for an input of 50,500 t/a woodchips and a wind park size of 36 MWp.

Coupled Thermo-Hydro-Geochemical Models of Engineered Barrier Systems: The Febex Project

2000 ◽  
Vol 663 ◽  
Author(s):  
J. Samper ◽  
R. Juncosa ◽  
V. Navarro ◽  
J. Delgado ◽  
L. Montenegro ◽  
...  
Keyword(s):  
Large Scale ◽  
Reference Model ◽  
Numerical Models ◽  
Full Scale ◽  
Small Scale ◽  
Model Parameters ◽  
In Situ Test ◽  
Wide Range ◽  
Engineered Barrier

ABSTRACTFEBEX (Full-scale Engineered Barrier EXperiment) is a demonstration and research project dealing with the bentonite engineered barrier designed for sealing and containment of waste in a high level radioactive waste repository (HLWR). It includes two main experiments: an situ full-scale test performed at Grimsel (GTS) and a mock-up test operating since February 1997 at CIEMAT facilities in Madrid (Spain) [1,2,3]. One of the objectives of FEBEX is the development and testing of conceptual and numerical models for the thermal, hydrodynamic, and geochemical (THG) processes expected to take place in engineered clay barriers. A significant improvement in coupled THG modeling of the clay barrier has been achieved both in terms of a better understanding of THG processes and more sophisticated THG computer codes. The ability of these models to reproduce the observed THG patterns in a wide range of THG conditions enhances the confidence in their prediction capabilities. Numerical THG models of heating and hydration experiments performed on small-scale lab cells provide excellent results for temperatures, water inflow and final water content in the cells [3]. Calculated concentrations at the end of the experiments reproduce most of the patterns of measured data. In general, the fit of concentrations of dissolved species is better than that of exchanged cations. These models were later used to simulate the evolution of the large-scale experiments (in situ and mock-up). Some thermo-hydrodynamic hypotheses and bentonite parameters were slightly revised during TH calibration of the mock-up test. The results of the reference model reproduce simultaneously the observed water inflows and bentonite temperatures and relative humidities. Although the model is highly sensitive to one-at-a-time variations in model parameters, the possibility of parameter combinations leading to similar fits cannot be precluded. The TH model of the “in situ” test is based on the same bentonite TH parameters and assumptions as for the “mock-up” test. Granite parameters were slightly modified during the calibration process in order to reproduce the observed thermal and hydrodynamic evolution. The reference model captures properly relative humidities and temperatures in the bentonite [3]. It also reproduces the observed spatial distribution of water pressures and temperatures in the granite. Once calibrated the TH aspects of the model, predictions of the THG evolution of both tests were performed. Data from the dismantling of the in situ test, which is planned for the summer of 2001, will provide a unique opportunity to test and validate current THG models of the EBS.

Model-Based Analysis of Transient Behavior of Large Scale CFB Boilers

2003 ◽  
Author(s):  
Ari Kettunen ◽  
Timo Hyppa¨nen ◽  
Ari-Pekka Kirkinen ◽  
Esa Maikkola
Keyword(s):  
Flow Rate ◽  
Large Scale ◽  
Air Flow ◽  
Model Parameters ◽  
Air Flow Rate ◽  
Process Data ◽  
Cfb Boiler ◽  
Flow Rates ◽  
Load Change ◽  
Secondary Air

The main objective of this study was to investigate the load change capability and effect of the individual control variables, such as fuel, primary air and secondary air flow rates, on the dynamics of large-scale CFB boilers. The dynamics of the CFB process were examined by dynamic process tests and by simulation studies. A multi-faceted set of transient process tests were performed at a commercial 235 MWe CFB unit. Fuel reactivity and interaction between gas flow rates, solid concentration profiles and heat transfer were studied by step changes of the following controllable variables: fuel feed rate, primary air flow rate, secondary air flow rate and primary to secondary air flow ratio. Load change performance was tested using two different types of tests: open and closed loop load changes. A tailored dynamic simulator for the CFB boiler was built and fine-tuned by determining the model parameters and by validating the models of each process component against measured process data of the transient test program. The know-how about the boiler dynamics obtained from the model analysis and the developed CFB simulator were utilized in designing the control systems of three new 262 MWe CFB units, which are now under construction. Further, the simulator was applied for the control system development and transient analysis of the supercritical OTU CFB boiler.

scholarly journals Are streamflow recession characteristics really characteristic?

2013 ◽  
Vol 17 (2) ◽  
pp. 817-828 ◽  
Author(s):  
M. Stoelzle ◽  
K. Stahl ◽  
M. Weiler
Keyword(s):  
Extraction Method ◽  
Storage Capacity ◽  
Extraction Methods ◽  
Model Parameters ◽  
Catchment Scale ◽  
Model Parameterization ◽  
Recession Curves ◽  
Recession Analysis ◽  
Meso Scale

Abstract. Streamflow recession has been investigated by a variety of methods, often involving the fit of a model to empirical recession plots to parameterize a non-linear storage–outflow relationship based on the dQ/dt−Q method. Such recession analysis methods (RAMs) are used to estimate hydraulic conductivity, storage capacity, or aquifer thickness and to model streamflow recession curves for regionalization and prediction at the catchment scale. Numerous RAMs have been published, but little is known about how comparably the resulting recession models distinguish characteristic catchment behavior. In this study we combined three established recession extraction methods with three different parameter-fitting methods to the power-law storage–outflow model to compare the range of recession characteristics that result from the application of these different RAMs. Resulting recession characteristics including recession time and corresponding storage depletion were evaluated for 20 meso-scale catchments in Germany. We found plausible ranges for model parameterization; however, calculated recession characteristics varied over two orders of magnitude. While recession characteristics of the 20 catchments derived with the different methods correlate strongly, particularly for the RAMs that use the same extraction method, not all rank the catchments consistently, and the differences among some of the methods are larger than among the catchments. To elucidate this variability we discuss the ambiguous roles of recession extraction procedures and the parameterization of the storage–outflow model and the limitations of the presented recession plots. The results suggest strong limitations to the comparability of recession characteristics derived with different methods, not only in the model parameters but also in the relative characterization of different catchments. A multiple-methods approach to investigating streamflow recession characteristics should be considered for applications whenever possible.

A Micromechanics Approach to Assess Ductile Crack Initiation in Damaged Pipelines

2003 ◽  
Author(s):  
Claudio Ruggieri ◽  
Fernando F. Santos ◽  
Mitsuru Ohata ◽  
Masao Toyoda
Keyword(s):  
Ductile Fracture ◽  
Large Scale ◽  
Cell Model ◽  
High Strength ◽  
Void Coalescence ◽  
Model Parameters ◽  
Ductile Crack ◽  
Cracking Behavior ◽  
Damage Criterion ◽  
Limited Applicability

This study explores the capabilities of a computational cell framework into a 3-D setting to model ductile fracture behavior in tensile specimens and damaged pipelines. The cell methodology provides a convenient approach for ductile crack extension suitable for large scale numerical analyses which includes a damage criterion and a microstructural length scale over which damage occurs. Laboratory testing of a high strength structural steel provides the experimental stress-strain data for round bar and circumferentially notched tensile specimens to calibrate the cell model parameters for the material. The present work applies the cell methodology using two damage criterion to describe ductile fracture in tensile specimens: (1) the Gurson-Tvergaard (GT) constitutive model for the softening of material and (2) the stress-modified, critical strain (SMCS) criterion for void coalescence. These damage criteria are then applied to predict ductile cracking for a pipe specimen tested under cycling bend loading. While the methodology still appears to have limited applicability to predict ductile cracking behavior in pipe specimens, the cell model predictions of the ductile response for the tensile specimens show good agreemeent with experimental measurements.

Modeling and Identification of Elastomeric Bearing Characteristics for Large-Scale Simulations

1998 ◽  
Vol 120 (1) ◽  
pp. 63-73 ◽  
Author(s):  
K. N. Morman ◽  
E. Nikolaidis ◽  
J. Rakowska ◽  
S. Seth
Keyword(s):  
Large Scale ◽  
Dynamic Stiffness ◽  
Testing Procedure ◽  
Nonlinear Dependence ◽  
Model Parameters ◽  
Elastomeric Bearing ◽  
Nonlinear Viscoelastic ◽  
Component Test ◽  
Differential Type ◽  
Stiffness And Damping

A constitutive equation of the differential type is introduced to model the nonlinear viscoelastic response behavior of elastomeric bearings in large-scale system simulations for vibration assessment and component loads prediction. The model accounts for the nonlinear dependence of dynamic stiffness and damping on vibration amplitude commonly observed in the behavior of bearings made of particle-reinforced elastomers. A testing procedure for the identification of the model parameters from bearing component test data is described. The experimental and analytical results for predicting the behavior of four (4) different car bushings are presented. In an example application, the model is incorporated in an ADAMS simulation to study the dynamic behavior of a car rear suspension.

scholarly journals The Magnetized Disk-Halo Transition Region of M51

2018 ◽  
Vol 14 (A30) ◽  
pp. 319-322 ◽  
Author(s):  
M. Kierdorf ◽  
S. A. Mao ◽  
A. Fletcher ◽  
R. Beck ◽  
M. Haverkorn ◽  
...  
Keyword(s):  
Transition Region ◽  
Large Scale ◽  
Model Parameters ◽  
Very Large Array ◽  
X Band ◽  
Model Predictions ◽  
The Galaxy ◽  
Linearly Polarized ◽  
Grand Design ◽  
Different Depths

AbstractAn excellent laboratory for studying large scale magnetic fields is the grand design face-on spiral galaxy M51. Due to wavelength-dependent Faraday depolarization, linearly polarized synchrotron emission at different radio frequencies gives a picture of the galaxy at different depths: Observations at L-band (1 – 2 GHz) probe the halo region while at C- and X-band (4 – 8 GHz) the linearly polarized emission probe the disk region of M51. We present new observations of M51 using the Karl G. Jansky Very Large Array (VLA) at S-band (2 – 4 GHz), where previously no polarization observations existed, to shed new light on the transition region between the disk and the halo. We discuss a model of the depolarization of synchrotron radiation in a multilayer magneto-ionic medium and compare the model predictions to the multi-frequency polarization data of M51 between 1 – 8 GHz. The new S-band data are essential to distinguish between different models. Our study shows that the initial model parameters, i.e. the total regular and turbulent magnetic field strengths in the disk and halo of M51, need to be adjusted to successfully fit the models to the data.

scholarly journals Self-Supervised Pre-Training of Transformers for Satellite Image Time Series Classification

2020 ◽  
Author(s):  
Yuan Yuan ◽  
Lei Lin
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Large Scale ◽  
Temporal Structure ◽  
Satellite Image ◽  
Fine Tuning ◽  
Small Scale ◽  
Model Parameters ◽  
Learning Approaches ◽  
Wide Range

Satellite image time series (SITS) classification is a major research topic in remote sensing and is relevant for a wide range of applications. Deep learning approaches have been commonly employed for SITS classification and have provided state-of-the-art performance. However, deep learning methods suffer from overfitting when labeled data is scarce. To address this problem, we propose a novel self-supervised pre-training scheme to initialize a Transformer-based network by utilizing large-scale unlabeled data. In detail, the model is asked to predict randomly contaminated observations given an entire time series of a pixel. The main idea of our proposal is to leverage the inherent temporal structure of satellite time series to learn general-purpose spectral-temporal representations related to land cover semantics. Once pre-training is completed, the pre-trained network can be further adapted to various SITS classification tasks by fine-tuning all the model parameters on small-scale task-related labeled data. In this way, the general knowledge and representations about SITS can be transferred to a label-scarce task, thereby improving the generalization performance of the model as well as reducing the risk of overfitting. Comprehensive experiments have been carried out on three benchmark datasets over large study areas. Experimental results demonstrate the effectiveness of the proposed method, leading to a classification accuracy increment up to 1.91% to 6.69%. <div><b>This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.</b></div>

Sign in / Sign up

Export Citation Format

Share Document