Computational Efficiency Research Articles

EB-Net: an efficient balanced network for accuracy and speed of remote sensing detection

ABSTRACT Remote sensing detection is a difficult task that requires not only identifying objects with complex background, quality, and angle issues, but also being lightweight enough to be carried by edge devices. In real remote sensing scenarios, achieving accurate, fast, and low-resource-consumption automated detection remains a significant challenge. Therefore, this paper proposes an efficient balanced network (EB-Net) for real remote sensing devices. First, a dynamic sparse attention (DSA) mechanism is proposed and has been proven with high performance via the complexity analysis. In addition, a new dynamic sparse transformer (DSFormer) is constructed using DSA, which enhances feature information and adapts to image resolution by self-attention and multi-headed attention, and achieves more flexible computation by random sampling. Then, three versions of discrete distribution IoU (DDIoU) are defined for adapting various scenarios and tasks in remote sensing, and this loss function makes the model achieve high accuracy and more lightweight. Finally, to make the model more lightweight, a cropped SPPF (CroSPPF) is presented, which significantly improves the computational efficiency by the lightweight of the sequence structure and activation function. Ablation experiments are conducted on the NWPU VHR-10 dataset, and demonstrate the effectiveness of the proposed methods. Numerous comparisons with state-of-the-art detectors are conducted on the NWPU VHR-10, RSOD and DOTA datasets. The experiments show that EB-Net outperforms state-of-the-art remote sensing detection models in comprehensive performance and achieves end-to-end accurate and lightweight detection.

TorchMD-Net 2.0: Fast Neural Network Potentials for Molecular Simulations.

Achieving a balance between computational speed, prediction accuracy, and universal applicability in molecular simulations has been a persistent challenge. This paper presents substantial advancements in TorchMD-Net software, a pivotal step forward in the shift from conventional force fields to neural network-based potentials. The evolution of TorchMD-Net into a more comprehensive and versatile framework is highlighted, incorporating cutting-edge architectures such as TensorNet. This transformation is achieved through a modular design approach, encouraging customized applications within the scientific community. The most notable enhancement is a significant improvement in computational efficiency, achieving a very remarkable acceleration in the computation of energy and forces for TensorNet models, with performance gains ranging from 2× to 10× over previous, nonoptimized, iterations. Other enhancements include highly optimized neighbor search algorithms that support periodic boundary conditions and smooth integration with existing molecular dynamics frameworks. Additionally, the updated version introduces the capability to integrate physical priors, further enriching its application spectrum and utility in research. The software is available at https://github.com/torchmd/torchmd-net.

LLMEffiChecker:Understanding and Testing Efficiency Degradation of Large Language Models

Large Language Models (LLMs) have received much recent attention due to their human-level accuracy. While existing works mostly focus on either improving accuracy or testing accuracy robustness, the computation efficiency of LLMs, which is of paramount importance due to often vast generation demands and real-time requirements, has surprisingly received little attention. In this paper, we make the first attempt to understand and test potential computation efficiency robustness in state-of-the-art LLMs. By analyzing the working mechanism and implementation of 20,543 public-accessible LLMs, we observe a fundamental property in LLMs that could be manipulated in an adversarial manner to reduce computation efficiency significantly. Our interesting observation is that the output length determines the computation efficiency of LLMs instead of the input, where the output length depends on two factors: an often sufficiently large yet pessimistic pre-configured threshold controlling the max number of iterations and a runtime generated end of sentence (EOS) token. Our key motivation is to generate test inputs that could sufficiently delay the generation of EOS such that LLMs would have to go through enough iterations to satisfy the pre-configured threshold. We present LLMEffiChecker , which can work under both white-box setting and black-box setting. In the white-box scenario, LLMEffiChecker develops a gradient-guided technique that searches for a minimal and unnoticeable perturbation at character-level, token-level, and structure-level. In the black-box scenario, LLMEffiChecker employs a causal inference-based approach to find critical tokens and similarly applies three levels of imperceptible perturbation to them. Both the white-box and black-box settings effectively delay the appearance of EOS, compelling these inputs to reach the naturally-unreachable threshold. To demonstrate the effectiveness of LLMEffiChecker , we conduct a systematic evaluation on nine public-available LLMs: Google T5, AllenAI WMT14, Helsinki-NLP translator, Facebook FairSeq, UNICAMP-DL translator, MarianMT, Google FLAN-T5, MBZUAI LaMini-GPT and Salesforce CodeGen. Experimental results show that LLMEffiChecker can increase on average LLMs’ response latency and energy consumption by 325% to 3244% and 344% to 3616%, respectively, by perturbing just one character or token in the input sentence. Our case study shows that inputs generated by LLMEffiChecker significantly affect the battery power in real-world mobile devices ( i.e. , drain more than 30 times battery power than normal inputs).

Semi-TSGAN: Semi-Supervised Learning for Highlight Removal Based on Teacher-Student Generative Adversarial Network

Despite recent notable advancements in highlight image restoration techniques, the dearth of annotated data and the lightweight deployment of highlight removal networks pose significant impediments to further advancements in the field. In this paper, to the best of our knowledge, we first propose a semi-supervised learning paradigm for highlight removal, merging the fusion version of a teacher–student model and a generative adversarial network, featuring a lightweight network architecture. Initially, we establish a dependable repository to house optimal predictions as pseudo ground truth through empirical analyses guided by the most reliable No-Reference Image Quality Assessment (NR-IQA) method. This method serves to assess rigorously the quality of model predictions. Subsequently, addressing concerns regarding confirmation bias, we integrate contrastive regularization into the framework to curtail the risk of overfitting on inaccurate labels. Finally, we introduce a comprehensive feature aggregation module and an extensive attention mechanism within the generative network, considering a balance between network performance and computational efficiency. Our experimental evaluations encompass comprehensive assessments on both full-reference and non-reference highlight benchmarks. The results demonstrate conclusively the substantive quantitative and qualitative enhancements achieved by our proposed algorithm in comparison to state-of-the-art methodologies.

A rehearsal framework for computational efficiency in online continual learning

AbstractIn the realm of online continual learning, models are expected to adapt to an ever-changing environment. One of the most persistent hurdles in this adaptation is the mitigation of a phenomenon called "Catastrophic Forgetting" (CF). This critical condition occurs when models trained on non-identically distributed data lose performance in previously learned tasks. Rehearsal methods, leveraging the ability to replay older samples, aim to address this challenge by incorporating a buffer of past training samples. However, the absence of known task boundaries complicates the adaptation of current CF mitigation methods. This paper proposes a method attuned to data stream characteristics and online model performance in a resource-constrained environment. The number of training iterations and learning rate emerges as crucial hyperparameters, impacting the efficacy and efficiency of online continual learning. Up to this point, we propose a combination of Experience Replay methodologies, a Drift Detector, and various training convergence policies, specially tailored for scenarios with unknown task boundaries. Experimental results demonstrate the effectiveness of our approach, maintaining or enhancing performance compared to baseline methods, while significantly improving computational efficiency.

Lightweight Hash-Based Authentication Protocol for Smart Grids

Smart grids integrate information and communications technology into the processes of electricity production, transportation, and consumption, thereby enabling interactions between power suppliers and consumers to increase the efficiency of the power grid. To achieve this, smart meters (SMs) are installed in households or buildings to measure electricity usage and allow power suppliers or consumers to monitor and manage it in real time. However, SMs require a secure service to address malicious attacks during memory protection and communication processes and a lightweight communication protocol suitable for devices with computational and communication constraints. This paper proposes an authentication protocol based on a one-way hash function to address these issues. This protocol includes message authentication functions to address message tampering and uses a changing encryption key for secure communication during each transmission. The security and performance analysis of this protocol shows that it can address existing attacks and provides 105,281.67% better computational efficiency than previous methods.

Dynamic phase-differencing profilometry with number-theoretical phase unwrapping and interleaved projection

High-speed 3D measurement is receiving increasing attention. However, simultaneously achieving high computational efficiency, algorithmic robustness, and reconstructing ratio is challenging. Therefore, a dynamic phase-differencing profilometry (DPDP) is proposed. By capturing the minimum three phase-shifting sinusoidal deformed patterns and establishing a brand-new model, the phase difference between the object on the reference plane and the reference plane is directly resolved to effectively improve computational efficiency. Although it is wrapped, by using only two auxiliary complementary gratings with a purposely designed lower frequency, a DPDP-based number-theoretical temporal phase unwrapping (NT-TPU) algorithm is also proposed to unwrap the wrapped phase difference rather than the phase itself with high robustness. Furthermore, compared to existing PSP-based NT-TPU, the proposed NT-TPU can normally work under more relaxed restrictions. In order to accomplish a high reconstructing ratio, a pentabasic interleaved projection (PIP) strategy based on time division multiplexing is proposed. It can improve the reconstructing ratio from one reconstruction per every five patterns to an equivalent of one reconstruction per every 1.67 patterns. Experimental results demonstrate that the proposed method achieves high computational efficiency, high algorithmic robustness, and high reconstructing ratio simultaneously and has prospective application in high-speed 3D measurement.

Discrete Boltzmann model with split collision for nonequilibrium reactive flows

Abstract A multi-relaxation-time discrete Boltzmann model (DBM) with split collision is proposed for both subsonic and supersonic compressible reacting flows, where chemical reactions take place among various components. The physical model is based on a unified set of discrete Boltzmann equations that describes the evolution of each chemical species with adjustable acceleration, specific heat ratio, and Prandtl number. On the righ-hand side of discrete Boltzmann equations, the collision, force, and reaction terms denote the change rates of distribution functions due to self- and cross-collisions, external forces, and chemical reactions, respectively. The source terms can be calculated in three ways, among which the matrix inversion method possesses the highest physical accuracy and computational efficiency. Through Chapman-Enskog analysis, it is proved that the DBM is consistent with the reactive Navier-Stokes equations, Fick's law and Stefan-Maxwell diffusion equation in the hydrodynamic limit. Compared with the one-step-relaxation model, the split collision model offers a detailed and precise description of hydrodynamic, thermodynamic, and chemical nonequilibrium effects. Finally, the model is validated by six benchmarks, including multicomponent diffusion, mixture in the force field, Kelvin-Helmholtz instability, flame at constant pressure, opposing chemical reaction, and steady detonation.

The Alternating Direction Method of Multipliers for Sufficient Dimension Reduction

The minimum average variance estimation (MAVE) method has proven to be an effective approach to sufficient dimension reduction. In this study, we apply the computationally efficient optimization algorithm named alternating direction method of multipliers (ADMM) to a particular approach (MAVE or minimum average variance estimation) to the problem of sufficient dimension reduction (SDR). Under some assumptions, we prove that the iterative sequence generated by ADMM converges to some point of the associated augmented Lagrangian function. Moreover, that point is stationary. It also presents some numerical simulations on synthetic data to demonstrate the computational efficiency of the algorithm.

A fast bootstrap algorithm for causal inference with large data.

Estimating causal effects from large experimental and observational data has become increasingly prevalent in both industry and research. The bootstrap is an intuitive and powerful technique used to construct standard errors and confidence intervals of estimators. Its application however can be prohibitively demanding in settings involving large data. In addition, modern causal inference estimators based on machine learning and optimization techniques exacerbate the computational burden of the bootstrap. The bag of little bootstraps has been proposed in non-causal settings for large data but has not yet been applied to evaluate the properties of estimators of causal effects. In this article, we introduce a new bootstrap algorithm called causal bag of little bootstraps for causal inference with large data. The new algorithm significantly improves the computational efficiency of the traditional bootstrap while providing consistent estimates and desirable confidence interval coverage. We describe its properties, provide practical considerations, and evaluate the performance of the proposed algorithm in terms of bias, coverage of the true 95% confidence intervals, and computational time in a simulation study. We apply it in the evaluation of the effect of hormone therapy on the average time to coronary heart disease using a large observational data set from the Women's Health Initiative.

Quantum Vision Transformers for Quark–Gluon Classification

We introduce a hybrid quantum-classical vision transformer architecture, notable for its integration of variational quantum circuits within both the attention mechanism and the multi-layer perceptrons. The research addresses the critical challenge of computational efficiency and resource constraints in analyzing data from the upcoming High Luminosity Large Hadron Collider, presenting the architecture as a potential solution. In particular, we evaluate our method by applying the model to multi-detector jet images from CMS Open Data. The goal is to distinguish quark-initiated from gluon-initiated jets. We successfully train the quantum model and evaluate it via numerical simulations. Using this approach, we achieve classification performance almost on par with the one obtained with the completely classical architecture, considering a similar number of parameters.

A Hybrid Continuation Framework for Analyzing Nonlinear Normal Modes of Systems With Contact Nonlinearity

Abstract The prediction of the modal properties of structural systems with contact clearances and prestress presents a computational challenge, as the nonlinearity induced by piecewise-linear stiffness eliminates the use of efficient linear modal analysis techniques. The most common approach to obtaining the nonlinear normal modes (NNMs) of these structural systems is a numerical framework that integrates numerical integration, the shooting method, and the pseudo-arc-length continuation scheme. This numerical continuation framework (NCF) computes NNMs through iterative numerical calculations; thus, the computational cost of the nonlinear modal analysis of complex nonlinear systems, particularly piecewise-linear systems, becomes prohibitively expensive as the model size increases. In this study, a hybrid continuation framework (HCF) combining analytic and numerical methods is proposed to enable efficient computations of NNMs for systems with contact boundaries. This new hybrid framework utilizes a semi-analytic method to conduct the iterative shooting procedure; thus, the computational burden of the numerical continuation can be significantly reduced. The proposed method is demonstrated on spring-mass oscillators with contact elements, and the NNMs obtained using the proposed method are validated by those computed using the traditional numerical continuation framework. The modal properties of the systems can be computed using the proposed framework with significant speed-up. Furthermore, the modal properties, including internal resonance and sharp turning in NNM curves, of the piecewise-linear systems are identified and discussed.

Efficient optimization of an accelerator neutron source for neutron capture therapy using genetic algorithms.

In recent years, genetic algorithms have been applied in the field of nuclear technology design, producing superior optimization results compared to traditional methods. They can be employed in the design and optimization of beam shaping assemblies (BSA) BSA to obtain the desired neutron beams. But it should be noted that the direct combination of Monte Carlo methods with genetic algorithms requires a significant amount of computational resources and time. Design and optimize BSA more efficiently to achieve neutron beams that meet specified recommendations. We propose an approach of NSGA II with crucial variables which are identified by multivariate statistical techniques. This approach significantly reduces the problem sizes, thus reducing the time required for optimization. We illustrate this methodology using the example of BSA design for AB-BNCT. The computational efficiency has tripled with crucial variables. By using NSGA II, we obtained optimized models conforming to both the new and old version IAEA BNCT guidelines through a single optimization process and subjected them to phantom analysis. The results demonstrate that models obtained through this method can meet the IAEA recommendations with deep advantage depth (AD) and high absorbed ratio (AR). The genetic algorithm with crucial variables displays tremendous potential in addressing BSA optimization challenges.

A novel method for rapid estimation of active bone marrow dose for radiotherapy patients in epidemiological studies.

In a dedicated effort to improve the assessment of clonal hematopoiesis (CH) and study leukemia risk following radiotherapy, we are developing a large-scale cohort study among cancer patients who received radiation. To that end, it will be critical to analyze dosimetric parameters of red bone marrow (ABM) exposure in relation to CH and its progression to myeloid neoplasms, requiring reconstruction method for ABM doses of a large-scale patients rapidly and accurately. To support a large-scale cohort study on the assessment of clonal hematopoiesis and leukemia risk following radiotherapy, we present a new method for the rapid reconstruction of ABM doses of radiotherapy among cancer patients. The key idea of the presented method is to segment patient bones rapidly and automatically by matching a whole-body computational human phantom, in which the skeletal system is divided into 34 bone sites, to patient CT images via 3D skeletal registration. The automatic approach was used to segment site-specific bones for 40 radiotherapy patients. Also, we segmented the bones manually. The bones segmented both manually and automatically were then combined with the patient dose matrix calculated by the treatment planning system (TPS) to derive patient ABM dose. We evaluated the performance of the automatic method in geometric and dosimetric accuracy by comparison with the manual approach. The pelvis showed the best geometric performance [volume overlap fraction (VOF): 52% (mean) with 23% (σ) and average distance (AD): 0.8cm (mean) with 0.5cm (σ)]. The pelvis also showed the best dosimetry performance [absorbed dose difference (ADD): 0.7Gy (mean) with 1.0Gy (σ)]. Some bones showed unsatisfactory performances such as the cervical vertebrae [ADD: 5.2Gy (mean) with 10.8Gy (σ)]. This impact on the total ABM dose, however, was not significant. An excellent agreement for the total ABM dose was indeed observed [ADD: 0.4Gy (mean) with 0.4Gy (σ)]. The computation time required for dose calculation using our method was robust (about one minute per patient). We confirmed that our method estimates ABM doses across treatment sites accurately, while providing high computational efficiency. The method will be used to reconstruct patient-specific ABM doses for dose-response assessment in a large cohort study. The method can also be applied to prospective dose calculation within a clinical TPS to support clinical decision making at the point of care.

Inferring causal protein signalling networks from single‐cell data based on parallel discrete artificial bee colony algorithm

AbstractInferring causal protein signalling networks from human immune system cell data is a promising approach to unravel the underlying tissue signalling biology and dysfunction in diseased cells, which has attracted considerable attention within the bioinformatics field. Recently, Bayesian network (BN) techniques have gained significant popularity in inferring causal protein signalling networks from multiparameter single‐cell data. However, current BN methods may exhibit high computational complexity and ignore interactions among protein signalling molecules from different single cells. A novel BN method is presented for learning causal protein signalling networks based on parallel discrete artificial bee colony (PDABC), named PDABC. Specifically, PDABC is a score‐based BN method that utilises the parallel artificial bee colony to search for the global optimal causal protein signalling networks with the highest discrete K2 metric. The experimental results on several simulated datasets, as well as a previously published multi‐parameter fluorescence‐activated cell sorter dataset, indicate that PDABC surpasses the existing state‐of‐the‐art methods in terms of performance and computational efficiency.

Towards efficient control synthesis for nonlinear wave energy conversion systems: impedance-matching meets the spectral-domain

AbstractExisting studies within the literature that focus on designing parametric energy-maximizing controllers for Wave Energy Converter (WEC) systems predominantly rely on the impedance-matching (IM) principle, originally developed for linear time-invariant systems. Alternatively, iterative optimization routines are commonly employed for nonlinear WECs. However, these approaches often face a trade-off between effectiveness in maximizing energy extraction and computational efficiency. To address this limitation, this study proposes a computationally efficient controller tuning method for analogous synthesis in the case of nonlinear WECs. The proposed approach combines a statistical linearization technique known as spectral-domain modeling with the IM principle, to synthesize a Proportional–Integrative (PI) controller for a nonlinear WEC. Furthermore, a comparison is performed with two other synthesis methods: one based on a standard (i.e. linear) frequency-domain representation of the WEC that incorporates the IM principle, and the other employing a gradient-free optimization routine applied to the nonlinear time-domain model of the WEC for PI parameter tuning through exhaustive numerical search. A discussion on the effectiveness of each tuning method in maximizing energy absorption is provided, including an appraisal of their associated computational time requirements. Numerical analyses demonstrate that the proposed method, which integrates spectral-domain modeling and IM, can achieve (almost) optimal PI controller design for a nonlinear WEC. Furthermore, this study addresses the inaccuracies inherent in the frequency-domain approach and significantly reduces the computational time compared to the exhaustive search procedure. The findings of this research represent a significant advancement towards the development of simple, effective, and efficient IM-based techniques for synthesis of controllers in nonlinear WEC systems

Comparison of some Bayesian estimation methods for type-I generalized extreme value distribution with simulation

The Weibull distribution is considered one of the Type-I Generalized Extreme Value (GEV) distribution, and it plays a crucial role in modeling extreme events in various fields, such as hydrology, finance, and environmental sciences. Bayesian methods play a strong, decisive role in estimating the parameters of the GEV distribution due to their ability to incorporate prior knowledge and handle small sample sizes effectively. In this research, we compare several shrinkage Bayesian estimation methods based on the squared error and the linear exponential loss functions. They were adopted and compared by the Monte Carlo simulation method. The performance of these methods is assessed based on their accuracy and computational efficiency in estimating the scale parameter of the Weibull distribution. To evaluate their performance, we generate simulated datasets with different sample sizes and varying parameter values. A technique for pre-estimation shrinkage is suggested to enhance the precision of estimation. Simulation experiments proved that the Bayesian shrinkage estimator and shrinkage pre-estimation under the squared loss function method are better than the other methods because they give the least mean square error. Overall, our findings highlight the advantages of shrinkage Bayesian estimation methods for the proposed distribution. Researchers and practitioners in fields reliant on extreme value analysis can benefit from these findings when selecting appropriate Bayesian estimation techniques for modeling extreme events accurately and efficiently.

Multi-objective green vehicle scheduling problem considering time window and emission factors in ship block transportation

Logistics distribution is one of the main sources of carbon dioxide emissions at present, and there are also such distribution problems in the shipbuilding process. With the increasing attention paid to environmental problems, how to effectively reduce the energy consumption of block transportation and improve the utilization rate of resources in the factory is the key problem that China’s shipbuilding industry needs to solve at present. This article considers the time windows for block transportation tasks, as well as the self-loading constraints of different types of flat cars, and establishes an optimization model that minimizes the empty transport time and energy consumption of the flat cars as the optimization objective. Then, an Improved Genetic Whale Optimization Algorithm is designed, which combines the cross and mutation ideas of genetic algorithms and proposes a whale individual position updating mechanism under a mixed strategy. Furthermore, the performance and computational efficiency of the algorithm are verified through comparative analysis with other classical optimization algorithms on standard test examples. Finally, the shipyard’s block transportation example proves that the energy-saving ship block transportation scheduling method can effectively improve the efficiency of shipbuilding enterprise’s block transportation and reduce the energy consumption in the block transportation process. It proves the engineering practicality of the green dispatching method proposed in this paper, which can further provide a decision-making method for shipyard managers.

OPTIMIZATION OF CNN + MOBILENETV3 FOR INSECT IDENTIFICATION: TOWARD HIGH ACCURACY

Developments in the field of artificial intelligence and deep learning, particularly Convolutional Neural Networks (CNN) techniques, have expanded the research potential and applications in ecology, including efficient and accurate insect classification. However, there are challenges in achieving high levels of accuracy with similar computational efficiency. In response, the efficient MobileNetV3 architecture was investigated to improve the insect pest classification process. Through an analytical descriptive quantitative approach and insect datasets from Kaggle, this study tested the effectiveness of CNN models optimized with MobileNetV3. The results indicated that the optimized model achieved classification accuracy of up to 90%, with consistent performance between training and validation data and significant loss reduction. With high precision and processing efficiency, this discovery makes a substantial contribution to deep learning applications in the field of intelligent agriculture, promising methodological improvements for other classification problems. Despite offering a promising solution, this study recognizes the limitation in dataset diversity and suggests further exploration with more varied datasets to strengthen the model's application in actual agricultural practices.

Real-time and Intelligent Moving Targets Tracking based on UAV Remote Sensing Video Camera and Brain-like Computing Chips

Abstract. Moving target tracking technology based on Unmanned Aerial Vehicles (UAV) is widely used in many fields such as automatic inspection and emergency response. The existing moving target tracking methods usually have the problems of large computation and low tracking efficiency. Limited by the computing power of the UAV platform, real-time tracking and analysis of multiple targets based on the video data collected by UAV platform is a difficult task. In this paper, we proposed a novel Target Specific Filtering Tracking with Memory (TSFMTrack) method designed for UAV-based real-time tracking tasks, which involves a Tracklet Filtering Module (TFM) for capturing object appearance features and a Tracklet Matching Module (TMM) for bounding box association in each frame. By experimental comparison with other State-Of-The-Art (SOTA) methods on popular MOT and UAV tracking datasets, the TSFMTrack have shown obvious advantages in accuracy, computational efficiency and reliability. Furthermore, we deployed the TSFMTrack on the brain-inspired chip Lynchip KA200, the experimental results have shown that the TSFMTrack is effective on edge computational platform and suitable for UAV real-time tracking tasks.