Search

CuGBasis is a free and open-source CUDA®/Python library for efficient computation of scalar, vector, and matrix quantities crucial for the post-processing of electronic structure calculations. CuGBasis integrates high-performance Graphical Processing Unit (GPU) computing with the ease and flexibility of Python programming, making it compatible with a vast ecosystem of libraries. We showcase its utility as a Python library and demonstrate its seamless interoperability with existing Python software to gain chemical insight from quantum chemistry calculations. Leveraging GPU-accelerated code, cuGBasis exhibits remarkable performance, making it highly applicable to larger systems or large databases. Our benchmarks reveal a 100-fold performance gain compared to alternative software packages, including serial/multi-threaded Central Processing Unit and GPU implementations. This paper outlines various features and computational strategies that lead to cuGBasis's enhanced performance, guiding developers of GPU-accelerated code. [ABSTRACT FROM AUTHOR]

The design of heterogeneous catalysts generally involves optimizing the reactivity descriptor of adsorption energy, which is inevitably governed by the structure of surface-active sites. A prerequisite for understanding the structure–properties relationship is the precise identification of real surface-active site structures, rather than relying on conceived structures derived from bulk alloy properties. However, it remains a formidable challenge due to the dynamic nature of nanoalloys during catalytic reactions and the lack of accurate and efficient interatomic potentials for simulations. Herein, a generalizable deep-learning potential for the Ag–Pd–F system is developed based on a dataset encompassing the bulk, surface, nanocluster, amorphous, and point defected configurations with diverse compositions to achieve a comprehensive description of interatomic interactions, facilitating precise prediction of adsorption energy, surface energy, formation energy, and diffusion energy barrier and is utilized to investigate the structural evolutions of AgPd nanoalloys during fluorination. The structural evolutions involve the inward diffusion of F, the outward diffusion of Ag in Ag@Pd nanoalloys, the formation of surface AgFx species in mixed and Janus AgPd nanoalloys, and the shape deformation from cuboctahedron to sphere in Ag and Pd@Ag nanoalloys. Moreover, the effects of atomic diffusion and dislocation formation and migration on the reconstructing pathway of nanoalloys are highlighted. It is demonstrated that the stress relaxation upon F adsorption serves as the intrinsic driving factor governing the surface reconstruction of AgPd nanoalloys. [ABSTRACT FROM AUTHOR]

Expanding the pool of stable halide perovskites with attractive optoelectronic properties is crucial to addressing current limitations in their performance as photovoltaic (PV) absorbers. In this article, we demonstrate how a high-throughput density functional theory (DFT) dataset of halide perovskite alloys can be used to train accurate surrogate models for property prediction and subsequently perform inverse design using genetic algorithm (GA). Our dataset consists of decomposition energies, bandgaps, and photovoltaic efficiencies of nearly 800 pure and mixed composition ABX3 compounds from both the GGA-PBE and HSE06 functionals, and are combined with ∼100 experimental data points collected from the literature. Multi-fidelity random forest regression models are trained on the DFT + experimental dataset for each property using descriptors that one-hot encode composition, phase, and fidelity, and additionally include well-known elemental or molecular properties of species at the A, B, and X sites. Rigorously optimized models are deployed for experiment-level prediction over >150 000 hypothetical compounds, leading to thousands of promising materials with low decomposition energy, band gap between 1 and 2 eV, and efficiency of >15%. Surrogate models are further combined with GA using an objective function to maintain chemical feasibility, minimize decomposition energy, maximize PV efficiency, and keep bandgap between 1 and 2 eV; thus, hundreds more optimal compositions and phases are discovered. We present an analysis of the screened and inverse-designed materials, visualize ternary phase diagrams generated for many systems of interest using machine learning predictions, and suggest strategies for further improvement and expansion in the future. [ABSTRACT FROM AUTHOR]

Double-atom catalysts (DACs) for harmful nitrate (NO3−) electroreduction to valuable ammonia (eNO3RR) is attractive for both environmental remediation and energy transformation. However, the limited metal loading in most DACs largely hinders their applications in practical catalytic applications. Therefore, exploring ultrahigh-density (UHD) DACs with abundant active metal centers and excellent eNO3RR activity is highly desired under the site-distance effect. Herein, starting from the experimental M2N6 motif deposited on graphene, we firstly screened the low-density (LD) Mn2N6 and Fe2N6 DACs with high eNO3RR activity and then established an appropriate activity descriptor for the LD–DAC system. By utilizing this descriptor, the corresponding Mn2N6 and Fe2N6 UHD–DACs with dynamic, thermal, thermodynamic, and electrochemical stabilities, are identified to locate at the peak of activity volcano, exhibiting rather-low limiting potentials of −0.25 and −0.38 V, respectively. Further analysis in term of spin state and orbital interaction, confirms that the electronic state effect similar to that of LD–DACs enable the excellent eNO3RR activity to be maintained in the UHD–DACs. These findings highlight the promising application of Mn2N6 and Fe2N6 UHD–DACs in nitrate electroreduction for NH3 production and provide impetus for further experimental exploration of ultrahigh-density DACs based on their intrinsic electronic states. [ABSTRACT FROM AUTHOR]

We introduce ACEpotentials.jl, a Julia-language software package that constructs interatomic potentials from quantum mechanical reference data using the Atomic Cluster Expansion [R. Drautz, Phys. Rev. B 99, 014104 (2019)]. As the latter provides a complete description of atomic environments, including invariance to overall translation and rotation as well as permutation of like atoms, the resulting potentials are systematically improvable and data efficient. Furthermore, the descriptor's expressiveness enables use of a linear model, facilitating rapid evaluation and straightforward application of Bayesian techniques for active learning. We summarize the capabilities of ACEpotentials.jl and demonstrate its strengths (simplicity, interpretability, robustness, performance) on a selection of prototypical atomistic modelling workflows. [ABSTRACT FROM AUTHOR]

High-accuracy prediction of the physical properties of amorphous materials is challenging in condensed-matter physics. A promising method to achieve this is machine-learning potentials, which is an alternative to computationally demanding ab initio calculations. When applying machine-learning potentials, the construction of descriptors to represent atomic configurations is crucial. These descriptors should be invariant to symmetry operations. Handcrafted representations using a smooth overlap of atomic positions and graph neural networks (GNN) are examples of methods used for constructing symmetry-invariant descriptors. In this study, we propose a novel descriptor based on a persistence diagram (PD), a two-dimensional representation of persistent homology (PH). First, we demonstrated that the normalized two-dimensional histogram obtained from PD could predict the average energy per atom of amorphous carbon at various densities, even when using a simple model. Second, an analysis of the dimensional reduction results of the descriptor spaces revealed that PH can be used to construct descriptors with characteristics similar to those of a latent space in a GNN. These results indicate that PH is a promising method for constructing descriptors suitable for machine-learning potentials without hyperparameter tuning and deep-learning techniques. [ABSTRACT FROM AUTHOR]

Surface defect detection by machine vision has received increased attention concerning the quality control of universal joint bearings (UJB). The defect distribution and counting information are important for product quality optimisation. However, vision defect detection for UJB remains a challenging task due to the diversity of background textures and defect characters on multiple surfaces. In this study, a multi-surface defect detection (MSDD) method consisting of region segmentation, feature extraction and detection is proposed. First, defect regions are accurately localised by the proposed adaptive defect region segmentation algorithm, which suppresses the inference of background variety. Then, a novel defect feature named multimodal fusion shape descriptor that integrates the global information and local information of defects is constructed to generate the discriminative defect representation. Finally, a defect feature extraction ability transfer strategy based on the transfer learning mechanism is proposed to address the problem of insufficient defect samples. The experimental results show that our method achieves the best accuracy of 94.8% macro-F1 and processes 28 defect images per second. Besides, the application effect in the practical production line indicates that our method meets the accuracy and real-time requirements of MSDD for UJB. [ABSTRACT FROM AUTHOR]

With the exponential progress in the field of cheminformatics, the conventional modeling approaches have so far been to employ supervised and unsupervised machine learning (ML) and deep learning models, utilizing the standard molecular descriptors, which represent the structural, physicochemical, and electronic properties of a particular compound. Deviating from the conventional approach, in this investigation, we have employed the classification Read-Across Structure-Activity Relationship (c-RASAR), which involves the amalgamation of the concepts of classification-based quantitative structure-activity relationship (QSAR) and Read-Across to incorporate Read-Across-derived similarity and error-based descriptors into a statistical and machine learning modeling framework. ML models developed from these RASAR descriptors use similarity-based information from the close source neighbors of a particular query compound. We have employed different classification modeling algorithms on the selected QSAR and RASAR descriptors to develop predictive models for efficient prediction of query compounds' hepatotoxicity. The predictivity of each of these models was evaluated on a large number of test set compounds. The best-performing model was also used to screen a true external data set. The concepts of explainable AI (XAI) coupled with Read-Across were used to interpret the contributions of the RASAR descriptors in the best c-RASAR model and to explain the chemical diversity in the dataset. The application of various unsupervised dimensionality reduction techniques like t-SNE and UMAP and the supervised ARKA framework showed the usefulness of the RASAR descriptors over the selected QSAR descriptors in their ability to group similar compounds, enhancing the modelability of the dataset and efficiently identifying activity cliffs. Furthermore, the activity cliffs were also identified from Read-Across by observing the nature of compounds constituting the nearest neighbors for a particular query compound. On comparing our simple linear c-RASAR model with the previously reported models developed using the same dataset derived from the US FDA Orange Book (https://www.accessdata.fda.gov/scripts/cder/ob/index.cfm), it was observed that our model is simple, reproducible, transferable, and highly predictive. The performance of the LDA c-RASAR model on the true external set supersedes that of the previously reported work. Therefore, the present simple LDA c-RASAR model can efficiently be used to predict the hepatotoxicity of query chemicals. [ABSTRACT FROM AUTHOR]

This paper proposes a novel efficient descriptor for finger knuckle print (FKP) identification. Our proposed descriptor is called MDL (Magnitude, Direction, and Local patterns), as it describes the rich structure of FKP images by jointly encoding three types of discriminative information, namely gradient magnitude, direction, and local patterns. This joint representation allows obtaining reliable features as gradient values that are merged with local features could offer abundant information on the FKP structure and the edges contained therein. Note that MDL is strengthened by extracting the features in a block-wise manner to include the spatial relationship information. Furthermore, we consider a multimodal learning scheme in which our MDL and deep features are extracted from major and minor FKP modalities. For the sake of efficiency, we opt for the VGG-F deep network, which has a small depth, to extract the deep features. The VGG-F is well-suited to the nature of FKP images (i.e., lines and edges without complicated backgrounds), where networks with a few layers/parameters can efficiently learn the intrinsic characteristics of the FKP without suffering from the overfitting. The discriminant correlation analysis (DCA) algorithm is used to perform feature fusion. DCA has the advantage of efficiently and jointly performing features fusion and dimensionality reduction. We conduct comprehensive experiments on a public finger knuckle dataset. Experimental results demonstrate the effectiveness of the proposed method, which has outperformed several related methods. The obtained results confirm the complementarity of our MDL and deep features. Moreover, the proposed method has shown a high robustness to occlusion scenario, where some parts of the FKP image are missed. [ABSTRACT FROM AUTHOR]

Estimating the synthesizability is the prerequisite to discovering novel high‐entropy ceramics with exotic properties. Herein, combined with the high‐throughput experiments and machine learning (ML) methods, the synthesizability of high‐entropy hexaborides (HEB6) is investigated. To construct the database, 100 equimolar quinary HEB6 samples synthesized using a self‐developed high‐throughput solid‐state reaction technique and 20 potential synthesizability descriptors calculated from fundamental parameters of constituent precursors are simultaneously collected. By employing the ML and the genetic algorithms, an optimal model consisting of five synthesizability descriptors (l¯$\bar{l}$, δl${\delta }_l$, δχP${\delta }_{{\chi }^{\mathrm{P}}}$, δI1${\delta }_{{I}_1}$, and δEF${\delta }_{{E}_{\mathrm{F}}}$) is determined for predicting the synthesizability of equimolar HEB6 with a high validation accuracy (93.0%), and 7 005 new equimolar quinary HEB6 are then proposed. Moreover, the applicability of our established model on non‐equimolar HEB6 is explored by the prediction of 30 synthesizability diagrams of non‐equimolar quinary HEB6. A high accuracy of 90.9% is further validated by the synthesis experiments of 11 non‐equimolar HEB6 candidates. Our work establishes an effective ML model for assessing the synthesizability of both equimolar and non‐equimolar HEB6 and paves a promising way to accelerate the discovery of new synthetic accessible high‐entropy ceramics. [ABSTRACT FROM AUTHOR]

Lattice metamaterials, a class of structures utilized for geometric lightweighting and enhancing structural integrity, feature interconnected elements arranged in specific patterns. These patterns confer unique properties beneficial for improving stiffness, damping, energy absorption, and strength across various applications. However, computational simulation of lattice structures for design and optimization presents significant challenges due to nonlinear and elastic/inelastic effects like instabilities, contacts, rate-dependence, and plasticity. Current methods heavily rely on finite element analysis (FEA), yet they entail high computational complexity and do not fully align with experimental observations. Moreover, additive manufacturing (AM) technologies introduce additional computational complexity due to process-induced anisotropic behaviors. This paper proposes a computational model to construct an effective descriptor by integrating metamaterial topological information with AM conditions. This descriptor serves as a surrogate for FEA models. Additionally, a database is established to correlate the descriptor with experimentally acquired mechanical responses of additively manufactured metamaterials. The bidirectional operation of the envisaged descriptor fulfills two objectives: informing the mechanical response of novel metamaterial models and guiding design and AM processes based on desired outcomes. Model validation demonstrates significant concurrence between predicted and experimental results, evidencing the model's capability to capture inherent nonlinearity in both design and process. [ABSTRACT FROM AUTHOR]

This paper proposes a novel fault-tolerant control (FTC) scheme for real-time uncertainty estimation in nonlinear systems. It addresses the challenges arising from nonlinear dynamics in system inputs, states, and outputs, along with measurement uncertainties, within an output feedback framework. Our approach leverages two key components: 1) A neural network NN descriptor-based observer: this novel observer concurrently estimates both system states and sensor uncertainties. It is particularly capable of handling unbounded sensor uncertainties in specific situations. It utilizes NNs as universal approximators to capture the system's complex nonlinearities. 2) A robust model reference tracking controller: this controller employs the estimated states from the NN descriptor-based observer to achieve the desired system performance despite the existence of uncertainties. It exhibits robustness, guaranteeing system stability and asymptotic state tracking to a given reference model. The efficacy of the proposed FTC scheme is validated through theoretical analysis and its application to two real-world case studies. • Addressing systems with nonlinear inputs, states, and output dynamics, and modeling them using NN approximators. • Handling unknown and maybe potentially unbounded additive uncertainties. • Introducing a new model for all nonlinearities and sensor uncertainties and augments the system as a descriptor system. • Designing an observer to concurrently estimate the state of the original system and measurement uncertainties. • Using the MRC technique to handle stable and unstable systems with unknown parameters, ensuring the desired performance. [ABSTRACT FROM AUTHOR]

The combinatorial interpretation of the persistence diagram as a Möbius inversion was recently shown to be functorial. We employ this discovery to recast the Persistent Homology Transform of a geometric complex as a representation of a cellulation on $ \mathbb{S}^n $ to the category of combinatorial persistence diagrams. Detailed examples are provided. We hope this recasting of the PH transform will allow for the adoption of existing methods from algebraic and topological combinatorics to the study of shapes. [ABSTRACT FROM AUTHOR]

The development of non-noble metal electrocatalysts for the Oxygen Evolution Reaction (OER) is advancing towards the use of multi-element materials. To reveal the complex correlations of multi-element OER electrocatalysts, we developed an iterative workflow combining high-throughput experiments and AI-generated content (AIGC) processes. An increased number of 909 (compared to 145 in previous literature) universal descriptors for inorganic materials science were constructed and used as Artificial Neural Network (ANN) input. A large number of statistical ensembles with each ANN individual ensemble having a reduced number of descriptors were integrated with a new Hierarchical Neural Network (HNN) algorithm. This algorithm addresses the longstanding challenge of balancing overwhelming descriptor numbers with insufficient datasets in traditional ANN approaches to materials science problems. As a result, the combination of AIGC and experimental validation significantly enhanced prediction accuracy, increase the R2 values from 0.7 to 0.98 for Tafel slopes. [ABSTRACT FROM AUTHOR]

Descriptor systems are more complex than normal systems, which are modeled by differential equations. This paper derives stability and stabilization criteria for uncertain fractional descriptor systems with neutral-type delay. Through the Lyapunov–Krasovskii functional approach, conditions subject to time-varying delay and parametric uncertainty are formulated as linear matrix inequalities. Based on the established criteria, static state- and output-feedback control laws are designed to ensure regularity and impulse-free properties, together with robust stability of the closed-loop system under permissible uncertainties. Numerical examples illustrate the effectiveness of the control methods and show that the results depend on the range of variation in the delays and on the fractional order, leading to stability analysis results that are less conservative than those reported in the literature. [ABSTRACT FROM AUTHOR]

In this paper, a stability analysis and the controller improvement of T-S fuzzy Descriptor system are studied. Firstly, by making full use of the related theory of fuzzy affiliation function and combining the design method of fuzzy Lyapunov function with the method of inequality deflation, a stability condition with wider admissibility and less system conservatism is proposed. The advantage of this method is that it is not necessary to ensure that each fuzzy subsystem is progressively stable. We also maximise the boundary of the derivatives of the affiliation function mined. Secondly, a PDC controller and a Non-PDC controller are designed, and the deflation conditions for the linear matrix inequalities of the two controllers are constructed. Finally, some arithmetic simulations and practical examples are given to demonstrate the effectiveness of the method studied in this paper, and the results obtained are less conservative and have larger feasible domains than previous methods. [ABSTRACT FROM AUTHOR]

This article presents two novel mixed‐uncertainty state estimators for discrete‐time descriptor linear systems, namely linear time‐varying mixed‐uncertainty filter (LTVMF) and linear time‐invariant mixed‐uncertainty filter (LTIMF). The former is based on the minimum‐variance approach, from which quadratic and explicit (special case of quadratic) formulations are derived and addressed to LTV systems. Both formulations incorporate the knowledge of state linear constraints, such as equality (in the descriptor form) and inequality, to mitigate precision and accuracy issues related to initialization and evolution of the state estimates. The LTIMF algorithm is based on the mixed H2/H∞$$ {H}_2/{H}_{\infty } $$ criterion and addressed to LTI systems, with the low‐cost computation being its motivation. Both LTVMF and LTIMF algorithms solve state‐estimation problems in which the uncertainties are combined to yield the so‐called mixed‐uncertainty vector, which is composed by set‐bounded uncertainties, characterized by constrained zonotopes, and stochastic uncertainties, characterized by Gaussian random vectors. As mixed‐uncertainty vectors imply biobjective optimization problems, we innovatively present multiobjective arguments to justify the choice of the solution on the Pareto‐optimal front. In order to discuss the advantages and drawbacks, the state estimators are tested in two numerical examples. [ABSTRACT FROM AUTHOR]

In this paper, a novel joint unknown input observer (JUIO) is proposed for a class of descriptor systems. The unknown input (UI) to be estimated injects additively into both the state and output equations in a state space model. To the best of our knowledge, only a few contributions in existing work address this problem directly. To begin with, by introducing an auxiliary UI, the original system is transformed into a normal form in which the output is no longer affected by UI. In this way, the negative effect brought by the UI occurring in the output measurement is removed. An interval observer is developed to obtain upper and lower boundary estimates of the output of the reformulated system. After that, an algebraic relationship between the auxiliary UI and the states is established, and a UI reconstruction (UIR) method is developed. Based on the UIR, a JUIO comprising the UIR and a Luenberger-like state observer is developed to achieve asymptotic estimations of the UI and state simultaneously. Verifiable conditions for the existence of the proposed JUIO are given with respect to the original descriptor system. Finally, a simulation example is presented to verify the effectiveness of the proposed method. • The UI in the output equation is removed without increasing the system dimension. • An auxiliary UIR is developed through an algebraic relationship between UI and state. • The JUIO can generate asymptotic state estimation and auxiliary UI simultaneously. [ABSTRACT FROM AUTHOR]

Keyframe extraction can effectively help users quickly understand video content. Generally, keyframes should be representative of the video content and simultaneously be diverse to reduce redundancy. Aiming to find the features of frames and filter out representative frames of the video, we propose a method of keyframe recommendation based on feature intercross and fusion (KFRFIF). The method is inspired by the implied relations between keyframe-extraction problem and recommendation problem. First, we investigate the application of a recommendation framework to the keyframe extraction problem. Second, the architecture of the proposed KFRFIF is put forward. Then, an algorithm for extracting intra-frame image features based on the combination of multiple image descriptors is proposed. An algorithm for extracting inter-frame distance features based on the combination of multiple distance calculation methods is designed. Moreover, A recommendation model based on feature intercross and fusion is put forward. An ablation study is further performed to verify the effectiveness of the submodule. Ultimately, the experimental results on four datasets with five outstanding approaches indicate the superior performance of our approach. [ABSTRACT FROM AUTHOR]

The term tertiary education has gained prominence in recent policy debates in the UK [Morgan (2023). “Labour ‘would review whole landscape’ of tertiary education.” Times Higher Education, 28 November. https://www.timeshighereducation.com/news/labour-would-review-whole-landscape-tertiary-education]. This paper presents findings from a review of the use of the term in academic and policy papers in the UK over the last 80 years. The analysis reveals the role of international organisations in defining this term initially as a synonym for post-compulsory education before using it as a statistical measure which became a normative guide for many member states. More recently, the term tertiary education has been used as a descriptor of systems of educational governance and by the World Bank for specific systems of organisation. Using these definitions as a guide, the review demonstrates differences in government policy in the four nations of the United Kingdom: England, Northern Ireland, Scotland and Wales. Here there is evidence of periodic and partial policy change, but little sustained full engagement with this agenda. The review suggests that if future tertiary education policy initiatives are introduced they should cover all potential learners once they have left the compulsory phase of education and should be advanced by regional bodies with involvement from learners, employers and community representatives allocating funds from a single tertiary education budget. [ABSTRACT FROM AUTHOR]

In this study, the electronic density of states (DOSs) calculated with density functional theory (DFT) were analyzed by the machine‐learning techniques. More than 400 pure metal and bimetallic alloy systems were calculated with DFT, and obtained the surface DOSs and the CH3 adsorption energy (Ead). By fitting the Gaussian functions to the DOS, multiple descriptors, such as the Gaussian peak positions, heights, and widths were extracted. Several regression methods, such as the least absolute shrinkage of selection operator (LASSO), random‐forest, gradient‐boosting, and extra‐tree were used to find the relationship between these descriptors and the Ead. The results show that the energy position of the peaks in the d‐projected DOS is the most important descriptor, in agreement with the previously known d‐band center theory. It was also shown that the peak position in d‐projected DOS improves the regression model in addition to the d‐band center, since it reduces the regression error. [ABSTRACT FROM AUTHOR]

Summary: This paper studies the problem of sensor fault estimation for linear systems in the presence of unknown inputs in both input and output channels. Relying on matrix equation theory, the unknown inputs are first removed from the output channel. Second, an augmented descriptor system is constructed by treating both system state and sensor fault as (pseudo)‐state variables. Third, the augmented descriptor system is transformed into a regular one, followed by the elimination of the unknown input from the state equation. Finally, a reduced‐order prescribed‐time sensor fault estimator is obtained by using periodic delayed observers. This matrix equation based approach is then complemented by an alternative one, which relies on successive transformations of state, input and outputs. This approach sheds light on the adopted solvability conditions throughout the paper and on the role of different directions in the output space in handling unknown inputs, sensor faults and the core state estimation problem. Finally, two numerical examples are presented to illustrate the effectiveness of the proposed methodologies. [ABSTRACT FROM AUTHOR]

Whether from environmental and occupational hazards or from topical pharmaceuticals, the human skin comes into contact with various chemicals every day. In vivo experiments not only require large investments of both time and money, but in vivo experiments can also be unethical due to the need to intentionally or incidentally expose humans or animals to toxic chemicals. Comparatively, in vitro experiments offer ethical and financial advantages when combined with the opportunity to selectively choose chemicals for experimentation. With in vivo experimentation being so infeasible, many scientists have chosen to make their in vitro data available publicly. Using these data, a detailed database containing 73 chemicals was created with a robust set of descriptors to be used in connection with mathematical modeling to predict diffusion, permeability, and partition coefficients. This resulting database is tailored to be easily used in various coding languages. [ABSTRACT FROM AUTHOR]

The cross-dimensional matching of 2D images and 3D point clouds is an effective method by which to establish the spatial relationship between 2D and 3D space, which has potential applications in remote sensing and artificial intelligence (AI). In this paper, we propose a novel multi-task network, 2D3D-DescNet, to learn 2D and 3D local feature descriptors jointly and perform cross-dimensional matching of 2D image patches and 3D point cloud volumes. The 2D3D-DescNet contains two branches with which to learn 2D and 3D feature descriptors, respectively, and utilizes a shared decoder to generate the feature maps of 2D image patches and 3D point cloud volumes. Specifically, the generative adversarial network (GAN) strategy is embedded to distinguish the source of the generated feature maps, thereby facilitating the use of the learned 2D and 3D local feature descriptors for cross-dimensional retrieval. Meanwhile, a metric network is embedded to compute the similarity between the learned 2D and 3D local feature descriptors. Finally, we construct a 2D-3D consistent loss function to optimize the 2D3D-DescNet. In this paper, the cross-dimensional matching of 2D images and 3D point clouds is explored with the small object of the 3Dmatch dataset. Experimental results demonstrate that the 2D and 3D local feature descriptors jointly learned by 2D3D-DescNet are similar. In addition, in terms of 2D and 3D cross-dimensional retrieval and matching between 2D image patches and 3D point cloud volumes, the proposed 2D3D-DescNet significantly outperforms the current state-of-the-art approaches based on jointly learning 2D and 3D feature descriptors; the cross-dimensional retrieval at TOP1 on the 3DMatch dataset is improved by over 12%. [ABSTRACT FROM AUTHOR]

This paper studies the output regulation problem of time-varying descriptor systems and the problem of designing state feedback and dynamic measurement output feedback control laws which asymptotically achieves output regulation and disturbance rejection is considered. New regulator equations are proposed for time-varying descriptor systems in the form of differential-algebraic matrix equations. The unique solution of the proposed regulator equations is given as well. We prove that the output regulation problem of time-varying descriptor systems is solvable if and only if the given regulator equations are solvable. Based on the solution of the regulator equations, the state feedback and dynamic measurement output feedback control laws are designed to solve the output regulation problem. The work extends the existing results of output regulation problem for time-varying linear systems to the time-varying descriptor systems. Numerical examples are given to show the effectiveness of our methodology. [ABSTRACT FROM AUTHOR]

The minimizing Delta Debugging (DDMIN) was among the first algorithms designed to automate the task of reducing test cases. Its popularity is based on the characteristics that it works on any kind of input, without knowledge about the input structure. Several studies proved that smaller outputs can be produced faster with more advanced techniques (e.g., building a tree representation of the input and reducing that data structure); however, if the structure is unknown or changing frequently, maintaining the descriptors might not be resource‐efficient. Therefore, in this paper, we focus on the evaluation of the novel fixed‐point iteration of minimizing Delta Debugging (DDMIN*) on publicly available test suites related to software engineering. Our experiments show that DDMIN* can help reduce inputs further by 48.08% on average compared to DDMIN (using lines as the units of the reduction). Although the effectiveness of the algorithm improved, it comes with the cost of additional testing steps. This study shows how the characteristics of the input affect the results and when it pays off using DDMIN*. [ABSTRACT FROM AUTHOR]

Introduction: Aging in rural settings worldwide, from the perspective of cognition, physical function, and life purpose essential constructs for a prosperous old age, still needs comprehensive discussion. This systematic review protocol aims to highlight the prevalence of cognitive decline, physical functioning, and life purpose in older adults aging in rural community settings. Methods and analysis: We will include cross-sectional studies published until April 2023 found in 8 databases: Embase, MEDLINE, LILACS, PsycINFO, Scopus, SciELO, and Web of Science. Ryyan software will be used for the first selection, and the Observational Study Quality Evaluation (OSQE) will assess methodological quality and risk of bias. Primary analysis will involve titles and abstracts using MeSH descriptors such as "Physical functioning," "Cognition," "Cognitive function," "Life purpose," "Elderly," "Older," "Rural aging," "Rural population," "Communities, rural," "Distribution, rural spatial," "Medium communities," "Rural settlement," "Small community." If necessary, secondary analysis will include a complete reading of selected articles by two blinded reviewers, confirmed by a third person. Publication bias will be assessed using cross-sectional analytical study quality. Sensitivity analyses will identify manuscripts significantly influencing combined prevalence of endpoints. [ABSTRACT FROM AUTHOR]

This paper proposes a novel algorithm, termed Mixed Riccati–Lyapunov Balanced Truncation (MRLBT), tailored for order reduction of Linear Time-Invariant Continuous-Time Descriptor Systems (LTI-CTD), commonly encountered in electrical and electronic circuit modeling. The MRLBT approach synergistically combines the advantages of balanced truncation (BT) and positive-real balanced truncation (PRBT) techniques while mitigating their limitations. Unlike BT, which preserves stability but not passivity, and PRBT, which retains passivity at the expense of larger reduction errors, MRLBT ensures the preservation of both stability and passivity inherent in the original system. Additionally, MRLBT achieves reduced computational complexity and minimized order reduction errors compared to PRBT. The proposed algorithm transforms the system into an equivalent Mixed Riccati–Lyapunov Balanced form, enabling the construction of a reduced-order model that retains the critical physical properties. Theoretical analysis and proofs are provided, establishing an upper bound on the global order reduction error. The efficacy of MRLBT is demonstrated through a numerical example involving an RLC ladder network, showcasing its superior performance over BT and PRBT in terms of reduced errors in the time and frequency domains. [ABSTRACT FROM AUTHOR]

In a finite-dimensional space, we consider a linear stochastic differential equation in Itô form with a singular constant matrix on the left-hand side. Taking into account various economic applications of such equations, they are classified as Leontief type equations, since under some additional assumptions, a deterministic analog of the equation in question describes the famous Leontief input–output balance model taking into account reserves. In the literature, these systems are more often called differential–algebraic or descriptor systems. In general, to study this type of equations, one needs higher-order derivatives of the right-hand side. This means that one must consider derivatives of the Wiener process, which exist in the generalized sense. In the previous papers, these equations were studied using the technique of Nelson mean derivatives of random processes, whose description does not require generalized functions. It is well known that mean derivatives depend on the -algebra used to find them. In the present paper, the study of this equation is carried out using mean derivatives with respect to a new -algebra that was not considered in the previous papers. [ABSTRACT FROM AUTHOR]

Similarity search is a key operation in content-based multimedia retrieval (CBMR) applications. Online CBMR applications, which are the focus of this work, perform a large number of search operations on dynamic datasets, which are updated at run-time. Additionally, the rates of search and data insertion (updated) operations vary during the execution. Such applications that rely on similarity search are required to fulfill these demands while also offering low response times. Thus, it is common for the computing demands in such applications to exceed the processing power of a single computer, motivating the usage of large-scale compute systems. As such, we propose in this work a distributed memory parallelization of similarity search that addresses these challenges. Our solution employs the efficient Inverted File System with Asymmetric Distance Computation algorithm (IVFADC) as the baseline, which is extended to support dynamic datasets. A dynamic resource management algorithm, called Multi-Stream Adaptation (MS-ADAPT) is proposed. It allows run-time changes on resource assignment with the goal of reducing response times. We evaluate our solution with multiple data partitioning strategies using up to 160 compute nodes and a dataset with 344 billion multimedia descriptors. Our experiments demonstrate superlinear scalability and MS-ADAPT outperforms the best static approach (oracle) by improving the response times up to 32 × on high-load cases. [ABSTRACT FROM AUTHOR]

The relation between passive and positive real systems has been extensively studied in the literature. In this paper, we study their connection to the more recently used notion of port-Hamiltonian descriptor systems. It is well-known that port-Hamiltonian systems are passive and that passive systems are positive real. Hence it is studied under which assumptions the converse implications hold. Furthermore, the relationship between passivity and KYP inequalities is investigated. [ABSTRACT FROM AUTHOR]

Synchronous generator system is a complicated dynamic system for energy transmission, which plays an important role in modern industrial production. In this article, we propose some predictor-corrector methods and structure-preserving methods for a generator system based on the first benchmark model of subsynchronous resonance, among which the structure-preserving methods preserve a Dirac structure associated with the so-called port-Hamiltonian descriptor systems. To illustrate this, the simplified generator system in the form of index-1 differential-algebraic equations has been derived. Our analyses provide the global error estimates for a special class of structure-preserving methods called Gauss methods, which guarantee their superior performance over the PSCAD/EMTDC and the predictor-corrector methods in terms of computational stability. Numerical simulations are implemented to verify the effectiveness and advantages of our methods. [ABSTRACT FROM AUTHOR]

This paper deals with the problem of robust decentralized control of discrete-time nonlinear interconnected descriptor systems (IDS) under the influence of external disturbances. First, we utilize the Takagi–Sugeno fuzzy model (TSFM) to represent discrete-time nonlinear IDS. For controller design, a proportional-derivative (PD) feedback strategy is employed to formulate a decentralized fuzzy controller for the discrete-time Takagi–Sugeno IDS (DTTSIDS). Furthermore, robust and passivity constraints are incorporated into the controller design process to mitigate the effects of disturbances. Based on the Lyapunov function and free weight function methods, we can analyze the stability of the DTTSIDS. The proposed conditions are converted into Linear Matrix Inequality (LMI) conditions through Schur Complement technology so that we can solve the problem through MATLAB LMI-Toolbox. Finally, the effectiveness of the proposed methodology is demonstrated through three examples presented in the paper and a comparison with other existing studies is given in Example 3. [ABSTRACT FROM AUTHOR]

Empirical formulas were derived for the interface shape, mechanical properties, and electrical characteristics of accumulative roll bonded (ARB) Cu/Nb laminated materials, based on relevant literature data. These formulas were incorporated into a forward analysis model using finite element analysis, enabling the calculation of yield stress and conductivity from the spatial distribution of Cu/Nb two phases. By randomly varying the layer thickness and interface shape in the two-phase spatial distribution and conducting repeated forward analyses, a database linking microstructural descriptors with yield stress and conductivity was created. These microstructural descriptors include volume fraction, geometric features, topological features, spatial correlation functions, and persistent homology. The significance of each microstructural descriptor on yield stress and conductivity was quantified using machine learning techniques. The results revealed that the Cu volume fraction, layer thickness, and 0th Betti number are crucial for yield stress, while for conductivity, the Cu volume fraction has the strongest influence, followed by layer thickness and layer continuity. Based on these outcomes, the Pareto front for ARB Cu/Nb laminates in the strength-conductivity space was presented. [ABSTRACT FROM AUTHOR]

This study proposes a method for classifying economic activity descriptors to match Nomenclature of Economic Activities (NACE) codes, employing a blend of machine learning techniques and expert evaluation. By leveraging natural language processing (NLP) methods to vectorize activity descriptors and utilizing genetic algorithm (GA) optimization to fine-tune hyperparameters in multi-class classifiers like Naive Bayes, Decision Trees, Random Forests, and Multilayer Perceptrons, our aim is to boost the accuracy and reliability of an economic classification system. This system faces challenges due to the absence of precise target labels in the dataset. Hence, it is essential to initially check the accuracy of utilized methods based on expert evaluations using a small dataset before generalizing to a larger one. [ABSTRACT FROM AUTHOR]

Summary: In this work, two state estimation methods are proposed for a class of discrete‐time descriptor linear systems subject to bounded uncertainties. First, we propose set‐valued estimator algorithm using symmetric boxes to compute rigorous bounds of the system states. The observer gains are calculated using L∞$$ {L}_{\infty } $$ norm to attenuate the effects of the uncertainties and to improve the accuracy of the proposed estimator. Second, to obtain tighter state enclosures, zonotopic set computation are developed instead of interval analysis to design a new set‐valued state estimation algorithm. The performances of the proposed state estimation approaches are highlighted on different illustrative examples. [ABSTRACT FROM AUTHOR]

This paper surveys some of the most important recent works related to micro-expression analysis. It includes discussions on algorithms for spotting and recognizing micro-expressions, their performances, databases, and new feature descriptor frameworks. Here, we use a reverse chronological order to discuss the state-of-the-art in a work-specific fashion. We provide a questionnaire-based comparative study of the literature and discuss future directions of micro-expression analysis. We expect this survey to expose micro-expressions to those interested in pursuing research in this emerging field. [ABSTRACT FROM AUTHOR]

The extraction of paleographical features is an important task to study the identity of the text in the Historical Manuscripts. One of the major features is the identification of the writer or copyist. Many researchers have worked on an automated system for writer identification, and with the development of deep learning techniques many approaches have been proposed. Most of the previous studies have developed a multi-steps system, while very few of them performed an End-to-End approach. Most of the systems rely on a pre-processing step to prepare the data in order to facilitate recognition. This paper presents an End-to-End deep learning system for writer identification, tested on four different datasets: ICDAR19 and ICFHR20 (Latin datasets), KHATT and Balamand (Arabic datasets). The system is based on the Deep-TEN approach using a customized ResNet-50 network for features and local descriptor extraction with an integration of a NetVLAD end-layer to compute and encode the global descriptor. It was compared with our state-of-the-art system, winner of ICFHR20 HisFrag competition, and showed an interesting performance on all datasets without any pre-processing techniques. [ABSTRACT FROM AUTHOR]

Significant progress has been made in recent years in the use of AI and Machine Learning (ML) for catalyst discovery and optimisation. The effectiveness of ML and data science techniques was demonstrated in predicting and optimising enantioselectivity and regioselectivity in catalytic reactions through optimisation of the ligands, counterions and reaction conditions. Direct discovery of new catalysts/reactions is more difficult and requires efficient exploration of transition metal chemical space. A range of computational techniques for descriptor generation, ranging from molecular mechanics to DFT methods, have been successfully demonstrated, often in conjunction with ML to reduce computational cost associated with TS calculations. Complex aspects of catalytic reactions, such as solvent, temperature, etc., have also been successfully incorporated into the ML optimisation and discovery workflow. [ABSTRACT FROM AUTHOR]

Machine-learned computational chemistry has led to a paradoxical situation in which molecular properties can be accurately predicted, but they are difficult to interpret. Explainable AI (XAI) tools can be used to analyze complex models, but they are highly dependent on the AI technique and the origin of the reference data. Alternatively, interpretable real-space tools can be employed directly, but they are often expensive to compute. To address this dilemma between explainability and accuracy, we developed SchNet4AIM, a SchNet-based architecture capable of dealing with local one-body (atomic) and two-body (interatomic) descriptors. The performance of SchNet4AIM is tested by predicting a wide collection of real-space quantities ranging from atomic charges and delocalization indices to pairwise interaction energies. The accuracy and speed of SchNet4AIM breaks the bottleneck that has prevented the use of real-space chemical descriptors in complex systems. We show that the group delocalization indices, arising from our physically rigorous atomistic predictions, provide reliable indicators of supramolecular binding events, thus contributing to the development of Explainable Chemical Artificial Intelligence (XCAI) models. Chemical AI often behaves as a black box, providing accurate, but opaque, predictions. Here, the authors show that the synergy of cutting-edge ANNs with the rigor of Quantum Chemical Topology can result in Explainable Chemical AI (XCAI). [ABSTRACT FROM AUTHOR]

Extreme learning machine (ELM) is a fast and efficient classifier. Due to the inability to process descriptor-level features extracted from video sequences, the networks based on ELM cannot be directly used to recognize human actions. Encoding learning network (ELN) is proposed to solve this problem. The network is composed of feature encoding module and double similarity-constrained extreme learning machine (DS-ELM). In feature encoding module, the sparse mapping weight matrix is combined with pyramid pooling to generate representation-level features. DS-ELM is used to classify generated features. In order to utilize the similarity information between the features of each layer, different weight matrices in ELN are separately trained to improve the recognition ability. In the training of sparse mapping weight matrix, the auto-encoded dictionary and similarity constrained linear coding (SCLC) method are proposed to encode the desired output. The sparse mapping weight matrix is trained by using partial descriptor features and corresponding desired outputs. In the training of the classification weights, the ELM objective function is updated by similarity relationship between hidden layer features to derive the training formula of DS-ELM, which improves the classification performance while avoiding iterative training. To verify the feasibility of the ELN, experiments are conducted on Olympic Sports, UCF11, Hollywood2, UCF101, and Self-collection databases. Experimental results show that the proposed ELN is able to directly process descriptor features. And, the similarity information between the features of each layer can be further utilized by ELN to obtain excellent recognition performance compared with other improved methods based on ELM. [ABSTRACT FROM AUTHOR]

This article investigates the fault‐tolerant formation containment control of multi‐agent systems with actuator and sensor faults. First, adaptive descriptor observers are proposed for the followers to estimate system states, actuator faults, and sensor faults simultaneously. Then, distributed fault‐tolerant formation containment controllers are designed based on the estimated values and sufficient conditions for the formation containment control are identified using Lyapunov method in the time domain. Finally, the effectiveness and advantages of the proposed method are demonstrated through numerical simulations. [ABSTRACT FROM AUTHOR]

Purpose: This study assesses the Multilayer Perceptron (MLP) neural network, complemented by other Machine Learning techniques (CART, PCA), in predicting the antimicrobial activity of 140 newly designed imidazolium chlorides against Klebsiella pneumoniae before synthesis. Emphasis is on leveraging molecular properties for predictive analysis. Methods: Classification and regression decision trees (CART) identified the top 200 predictive molecular descriptors. Principal Component Analysis (PCA) reduced these descriptors to 5 components, retaining 99.57% of raw data information. Antimicrobial activity, categorized as high or low, was based on experimentally proven minimal inhibitory concentration (MIC), with a cut-point at MIC = 0.856 mol/L. A 12-fold cross-validation trained the MLP (architecture 5-12-2 with 5 Principal Components). Results: The MLP exhibited commendable performance, achieving almost 90% correct classifications across learning, validation, and test sets, outperforming models without PCA dimension reduction. Key metrics, including accuracy (0.907), sensitivity (0.905), specificity (0.909), and precision (0.891), were notably high. These results highlight the MLP model's efficacy with PCA as a high-quality classifier for determining antimicrobial activity. Conclusions: The study concludes that the MLP neural network, along with CART and PCA, is a robust tool for predicting the antimicrobial activity class of imidazolium chlorides against Klebsiella pneumoniae. CART and PCA, used in this study, allowed input variable reduction without significant information loss. High classification accuracy and associated metrics affirm the method's potential utility in pre-synthesis assessments, offering valuable insights for antimicrobial compound design. [ABSTRACT FROM AUTHOR]

In general, the construction of observers for nonlinear descriptor systems depends on the solvability of a linear matrix inequality involving system matrices, and it is based on the system's nonlinearity. Therefore, the type of nonlinearity present in the system heavily affects the observer design process. There are significant developments in the literature for observer design for descriptor systems with various types of nonlinearity. Motivated by this, the current work reviews the literature on observer design for nonlinear descriptor systems with an extensive discussion on the type of nonlinearities that are considered. Here, an analysis and the comparison on the most common nonlinearities is presented, providing a roadmap to all researchers in the field. Furthermore, less common nonlinearities have been identified, presenting under-explored areas within the literature, and can open new domains for future research. [ABSTRACT FROM AUTHOR]

A differential pursuit game in a stochastic descriptor linear system is analyzed. The dynamic of the system is described by Ito's stochastic differential algebraic equation. Solutions of the equation are presented by the stochastic formula of the variation of constants in terms of the initial data and control unit. Constraints on the support functionals of two sets defined by the behaviors of the pursuer and evader are used to obtain the game completion conditions. The method of resolving functions is applied to construct pursuer's control bringing the dynamic vector of the system to the terminal set. The results are illustrated by an example of a stochastic descriptor system that describes transients in a radio engineering filter with random disturbances in the form of white noise. [ABSTRACT FROM AUTHOR]

The aim of this paper is to discuss the effect of the sensor on the acoustic emission (AE) signature and to develop a methodology to reduce the sensor effect. Pencil leads are broken on PMMA plates at different source–sensor distances, and the resulting waves are detected with different sensors. Several transducers, commonly used for acoustic emission measurements, are compared with regard to their ability to reproduce the characteristic shapes of plate waves. Their consequences for AE descriptors are discussed. Their different responses show why similar test specimens and test conditions can yield disparate results. This sensor effect will furthermore make the classification of different AE sources more difficult. In this context, a specific procedure is proposed to reduce the sensor effect and to propose an efficient selection of descriptors for data merging. Principal Component Analysis has demonstrated that using the Z-score normalized descriptor data in conjunction with the Krustal–Wallis test and identifying the outliers can help reduce the sensor effect. This procedure leads to the selection of a common descriptor set with the same distribution for all sensors. These descriptors can be merged to create a library. This result opens up new outlooks for the generalization of acoustic emission signature libraries. This aspect is a key point for the development of a database for machine learning. [ABSTRACT FROM AUTHOR]

The objective of content-based image retrieval (CBIR) is to locate samples from a database that are akin to a query, relying on the content embedded within the images. A contemporary strategy involves calculating the similarity between compact vectors by encoding both the query and the database images as global descriptors. In this work, we propose an image retrieval method by using hierarchical K-means clustering to efficiently organize the image descriptors within the database, which aims to optimize the subsequent retrieval process. Then, we compute the similarity between the descriptor set within the leaf nodes and the query descriptor to rank them accordingly. Three tree search algorithms are presented to enable a trade-off between search accuracy and speed that allows for substantial gains at the expense of a slightly reduced retrieval accuracy. Our proposed method demonstrates enhancement in image retrieval speed when applied to the CLIP-based model, UNICOM, designed for category-level retrieval, as well as the CNN-based R-GeM model, tailored for particular object retrieval by validating its effectiveness across various domains and backbones. We achieve an 18-times speed improvement while preserving over 99% accuracy when applied to the In-Shop dataset, the largest dataset in the experiments. [ABSTRACT FROM AUTHOR]

Serial crystallography experiments at synchrotron and X‐ray free‐electron laser (XFEL) sources are producing crystallographic data sets of ever‐increasing volume. While these experiments have large data sets and high‐frame‐rate detectors (around 3520 frames per second), only a small percentage of the data are useful for downstream analysis. Thus, an efficient and real‐time data classification pipeline is essential to differentiate reliably between useful and non‐useful images, typically known as 'hit' and 'miss', respectively, and keep only hit images on disk for further analysis such as peak finding and indexing. While feature‐point extraction is a key component of modern approaches to image classification, existing approaches require computationally expensive patch preprocessing to handle perspective distortion. This paper proposes a pipeline to categorize the data, consisting of a real‐time feature extraction algorithm called modified and parallelized FAST (MP‐FAST), an image descriptor and a machine learning classifier. For parallelizing the primary operations of the proposed pipeline, central processing units, graphics processing units and field‐programmable gate arrays are implemented and their performances compared. Finally, MP‐FAST‐based image classification is evaluated using a multi‐layer perceptron on various data sets, including both synthetic and experimental data. This approach demonstrates superior performance compared with other feature extractors and classifiers. [ABSTRACT FROM AUTHOR]

In recent years, semantic segmentation has made significant progress in visual place recognition (VPR) by using semantic information that is relatively invariant to appearance and viewpoint, demonstrating great potential. However, in some extreme scenarios, there may be semantic occlusion and semantic sparsity, which can lead to confusion when relying solely on semantic information for localization. Therefore, this paper proposes a novel VPR framework that employs a coarse-to-fine image matching strategy, combining semantic and appearance information to improve algorithm performance. First, we construct SemLook global descriptors using semantic contours, which can preliminarily screen images to enhance the accuracy and real-time performance of the algorithm. Based on this, we introduce SemLook local descriptors for fine screening, combining robust appearance information extracted by deep learning with semantic information. These local descriptors can address issues such as semantic overlap and sparsity in urban environments, further improving the accuracy of the algorithm. Through this refined screening process, we can effectively handle the challenges of complex image matching in urban environments and obtain more accurate results. The performance of SemLook descriptors is evaluated on three public datasets (Extended-CMU Season, Robot-Car Seasons v2, and SYNTHIA) and compared with six state-of-the-art VPR algorithms (HOG, CoHOG, AlexNet_VPR, Region VLAD, Patch-NetVLAD, Forest). In the experimental comparison, considering both real-time performance and evaluation metrics, the SemLook descriptors are found to outperform the other six algorithms. Evaluation metrics include the area under the curve (AUC) based on the precision–recall curve, Recall@100%Precision, and Precision@100%Recall. On the Extended-CMU Season dataset, SemLook descriptors achieve a 100% AUC value, and on the SYNTHIA dataset, they achieve a 99% AUC value, demonstrating outstanding performance. The experimental results indicate that introducing global descriptors for initial screening and utilizing local descriptors combining both semantic and appearance information for precise matching can effectively address the issue of location recognition in scenarios with semantic ambiguity or sparsity. This algorithm enhances descriptor performance, making it more accurate and robust in scenes with variations in appearance and viewpoint. [ABSTRACT FROM AUTHOR]

The surge in remote sensing satellites and diverse imaging modes poses substantial challenges for ground systems. Swift and high-precision geolocation is the foundational requirement for subsequent remote sensing image applications. Breakthroughs in intelligent on-orbit processing now enable on-orbit geometric processing. In the absence of control data on board, a recent trend is to introduce reference data onto satellites. However, the pre-storage of traditional reference images or control point databases presents a significant challenge to the limited on-board data storage capacity. Therefore, oriented to the demand for control information acquisition during on-orbit geometry processing, we propose the construction of lightweight and stable feature databases. Initially, stable feature classes are obtained through iterative matching filtering, followed by re-extracting feature descriptors for each stable feature point location on the training images. Subsequently, the descriptors of each point location are clustered and fused using affinity propagation (AP) to eliminate redundancy. Finally, LDAHash is utilized to quantize floating-point descriptors into binary descriptors, further reducing the storage space. In our experiments, we utilize a variety of feature algorithms to assess the generality of our proposed method, thus extending the scope of the feature database and its applicability to various scenarios. This work plays a crucial role in advancing the technology of on-orbit geometry processing for remote sensing satellites. [ABSTRACT FROM AUTHOR]