Efficient Framework Research Articles

Faster-PGYOLO: an efficient framework for floating debris detection in inland waters

Faster-PGYOLO: an efficient framework for floating debris detection in inland waters

Computational Fluid Dynamics Modeling of Pressure-Retarded Osmosis: Towards a Virtual Lab for Osmotic-Driven Process Simulations

Pressure-Retarded Osmosis (PRO) is an osmotically driven membrane-based process that has recently garnered significant attention from researchers due to its potential for clean energy harvesting from salinity gradients. The complex interactions between mixed-mode channel flows and osmotic fluxes in real PRO membrane modules necessitate high-fidelity modeling approaches. In this work, an efficient CFD framework is developed for the 3D simulation of osmotically driven membrane processes. This approach is based on a two-way coupling between a CFD solver, which captures external concentration polarization (ECP) effects, and an analytical representation of internal concentration polarization (ICP). Consequently, the osmotic water flux and reverse salt flux (RSF) can be accurately determined, accounting for all CP effects without any limitations on the geometrical complexity of the membrane chamber or its flow mode/regime. The proposed model is validated against experimental data, showing good agreement across various PRO case studies. Additionally, the model’s flexibility to simulate other types of osmotically driven processes such as forward osmosis (FO) is examined. Thus, the contributions of ECP and ICP effects in local osmotic pressure drop along the membrane chamber are comprehensively compared for FO and PRO modes. Finally, the capability of the CFD model to simulate a lab-scale PRO module is demonstrated across a range of Reynolds numbers from low-speed laminar up to turbulent flow regimes.

An Optimized Method for Reconstruction of Transcriptional Regulatory Networks in Bacteria Using ChIP-exo and RNA-seq Datasets.

Transcriptional regulatory networks (TRNs) in bacteria are crucial for elucidating the mechanisms that regulate gene expression and cellular responses to environmental stimuli. These networks delineate the interactions between transcription factors (TFs) and their target genes, thereby uncovering the regulatory processes that modulate gene expression under varying environmental conditions. Analyzing TRNs offers valuable insights into bacterial adaptation, stress responses, and metabolic optimization from an evolutionary standpoint. Additionally, understanding TRNs can drive the development of novel antimicrobial therapies and the engineering of microbial strains for biofuel and bioproduct production. This protocol integrates advanced data analysis pipelines, including ChEAP, DEOCSU, and DESeq2, to analyze omics datasets that encompass genome-wide TF binding sites and transcriptome profiles derived from ChIP-exo and RNA-seq experiments. This approach minimizes both the time required and the risk of bias, making it accessible to non-expert users. Key steps in the protocol include preprocessing and peak calling from ChIP-exo data, differential expression analysis of RNA-seq data, and motif and regulon analysis. This method offers a comprehensive and efficient framework for TRN reconstruction across various bacterial strains, enhancing both the accuracy and reliability of the analysis while providing valuable insights for basic and applied research.

An Efficient Framework for Finding Similar Datasets Based on Ontology

Governments are embracing an open data philosophy and making their data freely available to the public to encourage innovation and increase transparency. However, the number of available datasets is still limited. Finding relationships between related datasets on different data portals enables users to search the relevant datasets. These datasets are generated from the training data, which need to be curated by the user query. However, relevant dataset retrieval is an expensive operation due to the preparation procedure for each dataset. Moreover, it requires a significant amount of space and time. In this study, we propose a novel framework to identify the relationships between datasets using structural information and semantic information for finding similar datasets. We propose an algorithm to generate the Concept Matrix (CM) and the Dataset Matrix (DM) from the concepts and the datasets, which is then used to curate semantically related datasets in response to the users’ submitted queries. Moreover, we employ the proposed compression, indexing, and caching algorithms in our proposed scheme to reduce the required storage and time while searching the related ranked list of the datasets. Through extensive evaluation, we conclude that the proposed scheme outperforms the existing schemes.

Interpreting Federated Learning (FL) Models on Edge Devices by Enhancing Model Explainability with Computational Geometry and Advanced Database Architectures

Federated learning (FL) on edge devices has emerged as a promising approach for decentralized model training, enabling data privacy and efficiency in distributed networks. However, the complexity of these models presents significant challenges in terms of transparency and interpretability, which are critical for trust and accountability in real-world applications. This paper explores the integration of explainable AI techniques to enhance model interpretability within federated learning systems. By incorporating computational geometry, we aim to optimize model structure and decision-making processes, providing clearer insights into how models generate predictions. Additionally, we examine the role of advanced database architectures in managing the complexity of federated learning models on edge devices, ensuring efficient data handling and storage. Together, these approaches contribute to a more transparent, efficient, and scalable framework for federated learning on edge networks, addressing key challenges in both model explainability and performance optimization. This review highlights recent advancements and suggests future directions for research at the intersection of federated learning (FL), edge computing, explainability, and computational techniques.

An Efficient and Stable Registration Framework for Large Point Clouds at Two Different Moments

Point cloud registration plays a great role in many application scenarios; however, the registration of large-scale point clouds for actual different moments suffers from the problems of low efficiency, low accuracy, and a lack of stability. In this paper, we propose a registration framework for large-scale point clouds at different moments, which firstly downsamples large-scale point clouds using a random sampling method, then performs a random expansion strategy to make up for the loss of information caused by the random sampling, then completes the first registration by a deep learning network based on the extraction of keypoints and feature descriptors in combination with RANSAC, and finally completes the registration using the point-to-point ICP method. We conducted validation experiments and application experiments on large-scale point clouds of key train components, and the experimental results are much higher in accuracy or efficiency than other methods, which proves the effectiveness of our framework, which can be applied to actual large-scale point clouds.

Big data management, digital maintenance services and data analytics on intelligent operation: Basis for data management efficiency framework

Big data management, digital maintenance services and data analytics on intelligent operation: Basis for data management efficiency framework

Efficient transformer tracking with adaptive attention

AbstractRecently, several trackers utilising Transformer architecture have shown significant performance improvement. However, the high computational cost of multi‐head attention, a core component in the Transformer, has limited real‐time running speed, which is crucial for tracking tasks. Additionally, the global mechanism of multi‐head attention makes it susceptible to distractors with similar semantic information to the target. To address these issues, the authors propose a novel adaptive attention that enhances features through the spatial sparse attention mechanism with less than 1/4 of the computational complexity of multi‐head attention. Our adaptive attention sets a perception range around each element in the feature map based on the target scale in the previous tracking result and adaptively searches for the information of interest. This allows the module to focus on the target region rather than background distractors. Based on adaptive attention, the authors build an efficient transformer tracking framework. It can perform deep interaction between search and template features to activate target information and aggregate multi‐level interaction features to enhance the representation ability. The evaluation results on seven benchmarks show that the authors’ tracker achieves outstanding performance with a speed of 43 fps and significant advantages in hard circumstances.

Resource use efficiency: Developing a generic framework for protein production systems and its application on dairy and fish

Resource use efficiency has become increasingly important as the global demand for natural resources continues to grow, necessitating innovative and sustainable utilization strategies. Despite the availability of tools to quantify resource use and environmental impacts, an integrated assessment of total resource use efficiency remains underdeveloped. Addressing this gap, particularly within the context of the protein transition from animal-based to plant-based sources, this work introduces a novel resource efficiency framework. This framework advances current environmental assessments by enabling the precise calculation of efficiency in protein production processes and process chains based on comprehensive resource utilization metrics. It uniquely incorporates the nature of process inputs, including their renewability and circularity, and supports both foreground and life cycle perspectives at the process and process chain levels. To demonstrate its applicability, two case studies—fish and dairy production systems—are evaluated using exergy and mass analyses.

Efficient paddle-wheel copper cluster-based metal-organic framework for catalytic conversion of CO2 to oxazolidinones under mild conditions

Efficient utilization of CO2 and its conversion into high-value-added products contribute to mitigating environmental issues such as the greenhouse effect. Among these methods, the carboxylation cyclization of CO2 with propargylic amine represents a green and atom-economical approach. Developing cost-effective catalyst that can promote this reaction under mild condition is desirable but challenging. Herein, a unique two-dimensional (2D) metal–organic framework (MOF) {[Cu(QCA)2]∙0.5DMF∙0.5H2O}n (1) (QCA =6-quinolinecarboxylic acid, DMF = N,N-dimethylformamide) have been synthesized and thoroughly characterized. Compound 1 was constructed from paddle-wheel dinuclear copper clusters [Cu2(COO)4] and carboxylate ligands, presenting a two-dimensional framework with new topology. Stability tests indicate that compound 1 undergoes reversible structural changes in different solvents and exhibits good thermal stability. 1 as heterogeneous catalyst enabled the transformation of CO2 and propargylic amine into oxazolidinones under mild conditions with broad substrate generality. In addition, 1 could maintain its catalytic performance with five continuous reactions. This work presents an infrequent example of a 2D noble metal free MOF-based catalyst that shows high catalytic efficiency under mild conditions.

Efficacy of the construction industry legislative framework in developing countries – an Eswatini case

PurposeThe regulation of the construction industry remains a significant obstacle to economic progress in developing nations. Therefore, this study aimed to assess the efficacy of the legislative framework in the construction industry, focusing on Eswatini. The construction industry plays a critical role in the upliftment of the economy and an effective and efficient legislative framework is essential to ensure that as a developing country, Eswatini derives maximum benefits from the sector.Design/methodology/approachA phenomenological research design was used to explore stakeholders’ perceptions about the purpose of the legislative framework in the Eswatini construction industry and to establish the effectiveness of the current legislative framework in fulfilling its mandate. The data were collected using semi-structured interviews with open-ended questions. A total of 45 participants occupying different positions within the construction industry were selected using purposive sampling. The data were analysed thematically.FindingsThe results reiterate that the legislative framework’s purpose is to regulate, control and guide sectoral operations, including enforcing compliance with statutes and payment of construction levies. However, the legislative framework does not effectively fulfil its purpose of addressing stakeholder needs. Overall, the findings indicate shortcomings and inefficiencies within the framework aimed at regulating the construction industry of Eswatini. The inefficiencies have a great potential of defeating the purpose of the entire framework.Research limitations/implicationsRegarding implications for future research, the current study has set a new pathway for legislative framework analysis and evaluation. Therefore, future research on the topic can build on the current study and compare the legislative frameworks across developing countries. The purpose would be to establish the similarities and differences, and to identify strengths and weaknesses, possibly leading to established principles of what works in a given context. Other studies on the topic could focus on benchmarking against developed countries on how they measure the efficacy of the legislative framework within their constriction industry.Practical implicationsBased on the preceding, these findings may apply to similar developing countries at a similar stage of construction industry development. However, the findings should be considered within the limitations of the study. The focus was on developing countries, particularly the Eswatini context. Each country presents a unique context and challenges, so the findings should be cautiously generalized.Originality/valueWhile other studies in different contexts focus on developing legislative frameworks, the current study evaluated an existing legislative framework using defined criteria. The current study made a unique contribution by streamlining the features of effective and ineffective legislative frameworks. Hitherto, such information was scattered in literature but has been consolidated and applied empirically. Thus, the study has set benchmarks for analysing and evaluating legislative frameworks in developing countries.

An Efficient Deep Learning Framework for Optimized Event Forecasting

There have been several catastrophic events that have impacted multiple economies and resulted in thousands of fatalities, and violence has generated a severe political and financial crisis. Multiple studies have been centered around the artificial intelligence (AI) and machine learning (ML) approaches that are most widely used in practice to detect or forecast violent activities. However, machine learning algorithms become less accurate in identifying and forecasting violent activity as data volume and complexity increase. For the prediction of future events, we propose a hybrid deep learning (DL)-based model that is composed of a convolutional neural network (CNN), long short-term memory (LSTM), and an attention layer to learn temporal features from the benchmark the Global Terrorism Database (GTD). The GTD is an internationally recognized database that includes around 190,000 violent events and occurrences worldwide from 1970 to 2020. We took into account two factors for this experimental work: the type of event and the type of object used. The LSTM model takes these complex feature extractions from the CNN first to determine the chronological link between data points, whereas the attention model is used for the time series prediction of an event. The results show that the proposed model achieved good accuracies for both cases—type of event and type of object—compared to benchmark studies using the same dataset (98.1% and 97.6%, respectively).

An efficient leakage power optimization framework based on reinforcement learning with graph neural network

Threshold voltage (Vth\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$V_{th}$$\\end{document}) assignment is convenient for leakage optimization due to the exponential relation between leakage power and Vth\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$V_{th}$$\\end{document} by swapping logic cells without routing effort. However, it poses great challenge in large scale circuit design as an NP-hard problem. Machine learning-based approaches have been proposed to solve this problem, aiming to achieve well tradeoff between leakage power reduction and runtime speed up without new induced timing violation. In this paper, a leakage power optimization framework based on reinforcement learning (RL) with graph neural network (GNN) is first-ever proposed to formulate Vth\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$V_{th}$$\\end{document} assignment as a RL process by learning timing and physical characteristics of each circuit instance with GNN. Multiple instances are selected in a non-overlapped manner for each RL action iteration to speed up convergence and decouple timing interdependence along circuit path, where the corresponding reward is carefully defined to tradeoff between leakage reduction and potential timing violation. The proposed framework was validated by the Opencores and IWLS 2005 benchmark circuits with TSMC 28 nm technology. Experimental results demonstrate that our work outperforms prior non-analytical and GNN-based methods with better leakage power optimization by additional 5% to 17% reduction, which is highly consistent with the commercial tool. When transferring the trained RL-based framework to unseen circuits, it achieves the roughly identical leakage optimization results as seen circuit and speed up the runtime by 5.7×\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ imes$$\\end{document} to 8.5×\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ imes$$\\end{document} compared with commercial tool.

Integrated Characterization of Expanding-Solvent Steam-Assisted Gravity Drainage (ES-SAGD) Processes by Using a Heat-Penetration Criterion within a Unified, Consistent, and Efficient Framework

Summary The hybrid solvent-steam injection [e.g., expanding-solvent steam-assisted gravity drainage (ES-SAGD)] is the most promising method to enhance heavy oil recovery; however, it is quite a challenge to reproduce the experimental measurements and in-situ observations because of the complicated multiphase flow behavior resulting from the coupled mass and heat transfer. In this work, an integrated technique has been developed and applied for the first time to dynamically and accurately characterize an ES-SAGD process within a unified, consistent, and efficient framework. By taking the competitive impact between heat energy and solvent dissolution, a generalized heat-penetration (HP) criterion has been derived and integrated with a numerical simulator to characterize the dynamics of solvent/steam chamber propagation conditioned to the production profiles during hybrid solvent-steam processes. This generalized HP criterion allows us to not only dynamically calculate temperature profiles beyond a solvent/steam chamber interface (SCI) but also accurately and pragmatically quantify mass and heat transfer inside the diluted oil drainage zone as well as the solvent/steam chamber. Also, comprehensive effects of the thermally sensitive co/countercurrent flows are examined with a series of multiphase relative permeabilities. Such an integrated technique has been successfully validated by reproducing the measured solvent/steam chambers in 3D physical ES-SAGD experiments. Good agreements between the simulated and measured production profiles (i.e., injection temperature, pressure, and flow rate) have been made throughout the entire production period. Not only have the measured solvent/steam chambers been reproduced but also sensitivity analyses have been performed to investigate the influences of multiphase flow behavior, solvent concentration, and grid dimension. It is found that the diffusion/dispersion coefficients and thermal properties are dependent on temperature and solvent concentrations, competitively affecting the calculated temperature distributions. Moreover, gas-liquid relative permeabilities can impose a significant impact on the SCI moving velocity as well as the oil drainage front. Such an integrated approach considerably reduces the simulation uncertainties and complexities, offering a straightforward and effective means of dynamically reproducing the observed solvent/steam chambers within a unified, consistent, and efficient framework.

OXFORD-UMAT: An efficient and versatile crystal plasticity framework

The crystal plasticity-based finite element method is widely used, as it allows complex microstructures to be simulated and allows direct comparison with experiments. This paper presents the OXFORD-UMAT for Abaqus®, a novel crystal plasticity code that is publicly available online for researchers interested in using crystal plasticity. The model is able to simulate a wide range of materials and incorporates two different solvers based on the solution of slip increments and Cauchy stress with variants of state update procedures including explicit, semi-implicit, and fully-implicit for computational efficiency that can be set by the user based on the application. Constitutive laws are available for a range of materials with single or multiple phases for slip, creep, strain hardening, and back stress. The model includes geometrically necessary dislocations that can be computed using finite element interpolation functions by four alternative methods, including the total form with and without a correction for the dislocation flux, a widely used rate form, and a slip-gradient formulation. In addition, the initial strengthening and subsequent softening seen in irradiated materials can also be simulated with the model. The analysis is available in 2D (plane stress and plane strain) and 3D, including linear and quadratic elements. Here we include full derivations of the key equations used in the code and then demonstrate the capability of the code by modeling single-crystal and large-scale polycrystal cases. Comparison of OXFORD-UMAT with other available crystal plasticity codes for Abaqus® reveals the efficiency of the proposed approach, with the backup solver offering greater versatility for handling convergence issues commonly found in practical applications.

A multi-task learning framework for aerodynamic computation of two-dimensional airfoils

Accurate and efficient prediction of airfoil aerodynamic coefficients is essential for improving aircraft performance. However, current research often encounters significant challenges in balancing accuracy with computational efficiency when predicting complex aerodynamic coefficients. In this paper, a Multi-Task Learning framework for Aerodynamic parameters Computation (MTL4AC) of two-dimensional (2D) airfoils is proposed. The MTL4AC processes two key subtasks: flow field prediction and pressure coefficient prediction. These two subtasks complement each other to reveal both global and local aerodynamic changes around the airfoil. The flow field prediction provides a coarse-grained global perspective, which focuses on the pressure and velocity variations on and around the airfoil surface. The pressure coefficient prediction offers a fine-grained local perspective, which concentrates on the pressure distribution on the airfoil surface to accurately calculate lift and drag coefficients. The MTL4AC demonstrated substantial improvements in the experiments conducted on the public dataset, achieving significant enhancements in accuracy and stability. This research contributes an accurate and efficient framework for aerodynamic computation, integrating geometric features and advanced multi-task learning techniques to achieve superior performance in predicting aerodynamic coefficients.

PII-GCNet: Lightweight Multi-Modal CNN Network for Efficient Crowd Counting and Localization in UAV RGB-T Images

With the increasing need for real-time crowd evaluation in military surveillance, public safety, and event crowd management, crowd counting using unmanned aerial vehicle (UAV) captured images has emerged as an essential research topic. While conventional RGB-based methods have achieved significant success, their performance is severely hampered in low-light environments due to poor visibility. Integrating thermal infrared (TIR) images can address this issue, but existing RGB-T crowd counting networks, which employ multi-stream architectures, tend to introduce computational redundancy and excessive parameters, rendering them impractical for UAV applications constrained by limited onboard resources. To overcome these challenges, this research introduces an innovative, compact RGB-T framework designed to minimize redundant feature processing and improve multi-modal representation. The proposed approach introduces a Partial Information Interaction Convolution (PIIConv) module to selectively minimize redundant feature computations and a Global Collaborative Fusion (GCFusion) module to improve multi-modal feature representation through spatial attention mechanisms. Empirical findings indicate that the introduced network attains competitive results on the DroneRGBT dataset while significantly reducing floating-point operations (FLOPs) and improving inference speed across various computing platforms. This study’s significance is in providing a computationally efficient framework for RGB-T crowd counting that balances accuracy and resource efficiency, making it ideal for real-time UAV deployment.

Multi-agent Deep Reinforcement Learning for cloud-based digital twins in power grid management

As the power industry becomes increasingly complex, Digital Twin (DT) technology has emerged as a crucial tool for enhancing grid resilience and operational efficiency by creating dynamic digital replicas of physical systems. These replicas enable accurate simulations and proactive management, but the vast amount of data generated by DT systems poses significant challenges for processing and analysis. Cloud computing offers a flexible solution by offloading these computational tasks to distributed resources, allowing for real-time analysis and scalable operations. However, this approach introduces complexities in task distribution and maintaining quality of service. Recent efforts have applied Deep Reinforcement Learning (DRL) to address these challenges, primarily using single-agent methods. However, these methods struggle with scalability and performance in increasingly complex cloud environments. To overcome these limitations, we propose an efficient task scheduling framework based on Multi-Agent Deep Q-Network (MADQN) principles, specifically designed to optimize both response times and operational costs. We provide a comprehensive design overview of our approach and conduct a thorough evaluation of its performance. The experimental results clearly indicate that our approach can considerably reduce response times amd lower operational costs compared to current methods, including state-of-the-art single-agent approaches.

Privacy preserving verifiable federated learning scheme using blockchain and homomorphic encryption

This paper introduces a novel Privacy-Preserving Verifiable Federated Learning (PPVFL) scheme that integrates blockchain technology and homomorphic encryption to address critical challenges in decentralized machine learning. The proposed scheme ensures data privacy, integrity, verifiability, robust security, and efficiency in collaborative learning environments, particularly in sensitive domains such as healthcare. By leveraging blockchain’s decentralized, immutable ledger and homomorphic encryption’s capability to perform computations on encrypted data, the model maintains the confidentiality of sensitive information throughout the learning process. The inclusion of Byzantine fault tolerance and Elliptic Curve Digital Signature Algorithm (ECDSA) further enhances the system’s security against malicious attacks and data tampering, while the optimization of computational processes ensures efficient model training and communication. The novelty of this work lies in the seamless integration of blockchain and homomorphic encryption within a federated learning framework, specifically tailored for post-quantum cryptography, a combination that has not been extensively explored in prior research. This research represents a significant advancement in secure and efficient federated learning, offering a promising solution for industries that prioritize data privacy, security, and trust in collaborative machine learning. The effectiveness, security, and efficiency of the PPVFL scheme were validated using the Glaucoma dataset. The proposed method outperformed baseline federated learning algorithms, achieving a Dice coefficient of 0.918 and a Hausdorff distance of 4.05 on Severe Glaucoma (SG) cases, compared to 0.905 and 5.27, respectively, with traditional FedAvg. Moreover, the integration of blockchain and homomorphic encryption ensured that data privacy was upheld without compromising model performance, while efficient computation and communication processes minimized latency and resource consumption. This study contributes a robust, privacy-preserving, secure, efficient, and verifiable federated learning framework that addresses the pressing need for secure and scalable data management in distributed machine learning environments.

A Scalable, Secure, and Efficient Framework for Sharing Electronic Health Records Using Permissioned Blockchain Technology

This paper presents a scalable, secure blockchain-based healthcare system architecture that efficiently manages large patient datasets. DHTs and Skip Lists enable efficient data access, while DPoS and PBFT facilitate parallel transaction processing. Adaptive filters, Radix Trees extended by Merkle Trees, and an immutable blockchain ledger secured by Tendermint consensus ensure data integrity and protection against evolving threats. Threshold Cryptography secures consensus participant selection, and Bulletproofs verify transactions, complying with healthcare regulations. ChaCha20, a symmetric stream cipher, encrypts sensitive data, enhancing performance across devices. ABAC manages access rights, ensuring fine-grained control over data accessibility. This architecture offers a comprehensive, efficient, and secure solution for healthcare data management in blockchain environments.