Applied Soft Computing最新文献

{"title":"An instance-oriented multi-source information fusion technique based on neighborhood granules","authors":"Xiao Zhang ,&nbsp;Jingjing Shen ,&nbsp;Jinhai Li ,&nbsp;Xia Liu","doi":"10.1016/j.asoc.2025.113483","DOIUrl":"10.1016/j.asoc.2025.113483","url":null,"abstract":"<div><div>With the rapid development of society and technology, human beings have increasingly diverse sources of data collection. Consequently, multi-source information fusion techniques, aiming to utilize various technologies to process, integrate and analyze data from different sources to obtain valuable information, have attracted significant attention. As a structured and hierarchical manner by processing and analyzing data via “granules”, granular computing has been extensively applied to multi-source information fusion. Given that different information sources may contain redundant instances and noise at different levels, it is crucial to select representative instances from multiple information sources based on granular computing. However, there exists little research on instance-oriented fusion based on granular computing. To fill this gap, we investigate the issue of instance-oriented fusion in multi-source neighborhood decision information systems in this paper. Specifically, considering both the distribution and decision information of the neighborhood of an instance, we firstly propose the concept of internal confidence to reflect the reliable degree of an instance in an information source. Secondly, the external confidence is presented to measure the reliable degrees of information sources by employing the overlap degree of the neighborhood granules in multiple information sources. Then, by combining the internal confidence and the external confidence, we put forward a confidence index for instances within an information source to select representative instances from multiple information sources. Furthermore, we present an instance-oriented multi-source information fusion algorithm based on neighborhood granules (IoMsIF). Finally, the performance of IoMsIF is assessed by numerical experiments. The experimental results show that IoMsIF achieves satisfactory performance.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113483"},"PeriodicalIF":7.2,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144513759","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Contrastive learning-guided hashing model for artwork image retrieval","authors":"Zhenyu Wang,&nbsp;Yingdong Yang,&nbsp;Fucheng Wu,&nbsp;Wenjia Li","doi":"10.1016/j.asoc.2025.113486","DOIUrl":"10.1016/j.asoc.2025.113486","url":null,"abstract":"<div><div>With the growing demand for spiritual enrichment, artwork retrieval has gained increasing popularity in computer vision. While current quick response (QR) code-based and deep learning-based methods enable real-time artwork recognition, their dependence on manual annotations limits adaptability and scalability. In this paper, we propose an unsupervised hashing-based artwork retrieval framework that requires no manual labeling and adapts to new classes. The system comprises two key components: a detection module and a retrieval module. The detection module processes user-captured artwork images by predicting polygonal bounding boxes and rectifying them through perspective transformation algorithms. The retrieval module is the focus of our research. We present a novel contrastive learning framework for artwork retrieval that integrates a convolutional neural network (CNN) based feature extractor and a hashing encoder–decoder structure. This architecture processes input images into both floating-point and hashing representations while maintaining a binary memory bank for efficient similarity matching. Two specialized loss functions facilitate adaptive hashing encoding, aligning floating-point and hashing features through an unsupervised learning process. We evaluate our framework on a self-built dataset containing over 24k images spanning traditional Chinese paintings, oil paintings, and Chinese chops. The samples are unlabeled and there is only one image of each artwork. Comparative experiments with state-of-the-art methods demonstrate our system’s superior effectiveness and strong potential for practical artwork retrieval applications.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113486"},"PeriodicalIF":7.2,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144480181","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A population-based simulated annealing approach with adaptive mutation operator for solving the discounted {0-1} knapsack problem","authors":"Juntao Zhao ,&nbsp;Xiaochuan Luo","doi":"10.1016/j.asoc.2025.113480","DOIUrl":"10.1016/j.asoc.2025.113480","url":null,"abstract":"<div><div>The discounted {0-1} knapsack problem extends the traditional knapsack problem by incorporating unique discount relationships among item groups, adding complexity to the selection process. It finds applications in supply chain optimization, resource allocation, and financial portfolio management. The objective is to maximize profit while adhering to capacity constraints. This paper presents a novel approach that integrates population-based simulated annealing with adaptive differential evolution to efficiently solve the problem. The proposed algorithm introduces an advanced greedy randomized initialization, multi-neighborhood local optimization within simulated annealing framework, and use two differential evolution mutation operators (DE/current-to-rand/1/bin and DE/current-to-best/1/bin) to enhance exploration and exploitation. A comprehensive two-stage repair and re-optimization strategy is employed to handle infeasible solutions. Extensive testing on two groups of 80 benchmark instances highlights the algorithm’s robustness and performance, effectively tackling the complexities of the studied problem.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113480"},"PeriodicalIF":7.2,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144522346","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Reconstructing models for approximation errors in Takagi–Sugeno fuzzy control under imperfect matching","authors":"Jie Yang,&nbsp;Shao-Yan Gai,&nbsp;Fei-Peng Da","doi":"10.1016/j.asoc.2025.113489","DOIUrl":"10.1016/j.asoc.2025.113489","url":null,"abstract":"<div><div>This study aims to tackle the issues associated with stability analysis in Takagi–Sugeno (TS) fuzzy systems. A novel modeling technique is proposed to incorporate the approximation error information of membership functions (MFs) into the stability conditions. First, the classical piecewise linear approximation method is employed to decompose the MFs into a linear model and an associated error model. Then, a reconstruction strategy is introduced to transform the error model into a new fuzzy model, which is subsequently used to enhance the stability analysis of TS fuzzy systems. Compared with methods that consider only the extremal values of the error function, the proposed approach leads to a multiplicative enhancement in the amount of exploitable error information. Furthermore, stochastic disturbances are introduced to evaluate the robustness of the system. Finally, the effectiveness and practicality of the proposed method are validated through two simulation examples.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113489"},"PeriodicalIF":7.2,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144522344","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cost optimization strategy for dependent task offloading in vehicular edge computing","authors":"Tan Deng,&nbsp;Shixue Li,&nbsp;Mingfeng Huang,&nbsp;Xiaoyong Tang,&nbsp;Ronghui Cao,&nbsp;Wenzheng Liu,&nbsp;Yanping Wang","doi":"10.1016/j.asoc.2025.113441","DOIUrl":"10.1016/j.asoc.2025.113441","url":null,"abstract":"<div><div>The geometrically increasing computational demands strain vehicular systems. Vehicular edge computing effectively alleviate this by dividing tasks into sub-modules and offloading these modules to edge servers. However, the interdependence among subtasks makes task offloading and resource allocation highly challenging. To address this issue, we propose a cost optimization strategy for dependent task offloading in vehicular edge computing networks. Specifically, the task offloading process is divided into two sub-problems: task offloading and resource allocation. First, we propose a Sequenced Quantization based on the Recurrent Neural Network (SQ-RNN) algorithm for offloading decisions. This algorithm uses environmental information as the input of the RNN to generate an optimal task offloading strategy. which is then quantified into multiple binary offloading actions through an order-preserving quantization method. Then, we propose a resource allocation method based on Computing Resource Blocks (CRBs), which divides server resources into blocks and assigns them to tasks with the principle of balancing resource allocation and reducing costs. Finally, extensive simulation experiments conducted on real-world datasets demonstrate that our approach reduces computing delay by 15.27% computing energy consumption by 9.93%, and cost by approximately 10.16% on average within the experimental bandwidth range, compared to the baseline algorithm. Moreover, as the number of subtasks increases, the optimization effect becomes more pronounced.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113441"},"PeriodicalIF":7.2,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144470734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Controllable diffusion models for hazardous construction site scene generation","authors":"XunLong Wang,&nbsp;JiangTao Ren","doi":"10.1016/j.asoc.2025.113446","DOIUrl":"10.1016/j.asoc.2025.113446","url":null,"abstract":"<div><div>Hazardous scene recognition is critical for construction site safety, but the low occurrence of such scenes in real environments leads to insufficient training data, limiting model development. Hazardous scene generation helps address data scarcity but involves complex background-object relationships and significant size differences, making precise layout control difficult. These characteristics make achieving precise layout control challenging for general-purpose hazardous scene generation models. To address these challenges, we propose a novel construction site hazardous scene generation framework based on large language and diffusion models, consisting of a two-stage generation process. In the first stage, we fine-tune the large language model (LLM) through context learning to serve as a text-based layout generator. In the second stage, we introduce a novel text-to-image diffusion model to guide the image generation process, ensuring that the generated image adheres to the scene layout produced in the first stage. Additionally, We propose two key modules, the Layout Enhancement Module and the Scale Fusion Module, to improve image quality and layout adherence. Comparative experiments show that our method generates superior scenes with stronger controllability and higher quality. Testing on a real-world dataset achieved a mAP score of 38.1%, improving model accuracy by 20.4% AP, 17.5% <span><math><mrow><mi>A</mi><msub><mrow><mi>P</mi></mrow><mrow><mn>50</mn></mrow></msub></mrow></math></span>, and 17.0% AR compared to models trained on real data, demonstrating our method’s effectiveness in hazardous construction site scene generation.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113446"},"PeriodicalIF":7.2,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144502354","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Structural health monitoring of building structures using novel acceleration-based signal-to-image technique and 2D convolutional neural networks","authors":"Kunal Bharali ,&nbsp;Manashi Saharia ,&nbsp;Moumita Roy ,&nbsp;Nirmalendu Debnath","doi":"10.1016/j.asoc.2025.113457","DOIUrl":"10.1016/j.asoc.2025.113457","url":null,"abstract":"<div><div>Structural Health Monitoring (SHM) is essential for detecting damage in structures like buildings, bridges, aircraft, etc., as they can deteriorate over time or suffer sudden failure during extreme events. In this realm, the application of deep learning (DL) techniques has been observed to become quite an integral part of global-level SHM using acceleration-based measurements by automatic monitoring in a timely and precise manner. In the present work, a novel strategy has been proposed by considering acceleration-based measurements in the form of computer vision (through the encoding of time series into images) and proposes two novel strategies (employing DL models). Specifically, two advanced encoding strategies have been introduced: (i) compressed time series recurrence plot (CTSRP), which captures temporal dependencies and patterns in vibration data, and (ii) time delay encoding (dotTDE), which encodes time-delay features to enhance the representation of subtle damage characteristics. Transforming a time series to an image can bring contrasting features that may not be present in the raw time series. Such images are then utilized as input to the 2D convolutional neural network (CNN) for determining the possible damages in structures. The proposed strategies are finally validated using two building structures (numerically simulated) i.e., a shear frame structure and the IASC-ASCE benchmark structure (incorporating multiple damage scenarios). During the comparison of classification accuracy, it has been observed that the CTSRP method has achieved 96.1% accuracy and the dotTDE method has achieved 100% accuracy for the ten-story shear frame structure. These results have shown an improvement ranging from 0.9% to 92.9% for CTSRP and 5.0% to 100% for dotTDE over the state-of-the-art approaches. Additionally, for the IASC-ASCE benchmark structure, the CTSRP method has achieved 94.8% accuracy, with improvement ranging from 0.4% to 72.1%, while the dotTDE method has achieved 100% accuracy with improvement ranging from 5.9% to 77.3%. Moreover, the analysis of the robustness and scalability of the investigated DL models has also been reported. In the future, the investigation may be conducted by focusing on enhancing the proposed DL models to improve their ability to detect unseen damage classes and adapt to structural variations. Furthermore, the models can be applied to different real-world structures for broader validation and applicability.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113457"},"PeriodicalIF":7.2,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144522347","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An empirical study on optimizing binary spiking neural networks for neuromorphic computing","authors":"Ping He,&nbsp;Rong Xiao,&nbsp;Chenwei Tang,&nbsp;Shudong Huang,&nbsp;Jiancheng Lv","doi":"10.1016/j.asoc.2025.113471","DOIUrl":"10.1016/j.asoc.2025.113471","url":null,"abstract":"<div><div>Spiking neural networks process information using asynchronous spikes between neurons, making them ideal for handling spatiotemporal data from neuromorphic sensors and have shown superior performance in low-latency and low-power computing applications. However, current neuromorphic computing faces challenges such as high synaptic memory usage and complex synapse calculations, with the training process optimization lacking a solid foundation. Here, we present a hardware-friendly weight-binarized spiking neural network to reduce storage needs, accelerate optimization, and enhance computational efficiency. During training, weight binarization is applied to reduce memory size and access drastically. Meanwhile, we employ a hybrid optimizer that combines the Adam method with stochastic gradient descent to address the convergence challenges that arise from gradient sparsity due to the use of binary weights. During inference, a simple shift-based batch normalization algorithm is introduced to achieve the effect equivalent to the computationally expensive BN with low accuracy loss. Then, we empirically identify and study the effectiveness of various ad-hoc techniques on neuromorphic recognition tasks as a case study, providing best practices for optimization. To the best of our knowledge, this is the first work using systematic comparisons to reveal how commonly employed tricks are effective for training binary spiking neural networks. The implementations will be open-sourced on GitHub.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113471"},"PeriodicalIF":7.2,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144557212","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An ensemble local search framework for population-based metaheuristic algorithms on single-objective optimization","authors":"Chunlei Li ,&nbsp;Libao Deng ,&nbsp;Wenyin Gong ,&nbsp;Liyan Qiao","doi":"10.1016/j.asoc.2025.113462","DOIUrl":"10.1016/j.asoc.2025.113462","url":null,"abstract":"<div><div>Population-based metaheuristic algorithms are recognized as potent tools for tackling complex optimization problems. However, they are often plagued by premature convergence, making them susceptible to getting trapped in local optima. To mitigate this issue, this paper introduces an Ensemble Local Search (ENLS) framework that can seamlessly integrate with various metaheuristic algorithms. In ENLS, an online detection mechanism is designed to identify the occurrence of premature convergence during the search process. Subsequently, three local search strategies are triggered to assist the following three subpopulations based on the characteristic of each one: (1) the orthogonal learning mechanism is applied to design an orthogonal crossover-based local search strategy for the superior subpopulation, focusing on refining solutions within a narrow region; (2) a dynamic Lévy flight-based local search strategy is developed for the medium subpopulation to enhance the population diversity by leveraging the long-term short-step Lévy random walking; (3) the inferior subpopulation employs an opposition-based local search, incorporating a modified opposition-based learning mechanism to explore a broader space between inferior solutions and their opposite positions. By integrating these three local search strategies, the ENLS framework can effectively balance exploration and exploitation, addressing the challenge of premature convergence. To validate the effectiveness of ENLS, comparative experiments are conducted using 30 benchmark problems from the IEEE CEC 2014 test suite and 20 real-world optimization problems. The experimental results confirm that the ENLS framework significantly enhances the optimization capabilities of the considered 12 metaheuristic algorithms without significantly increasing runtime complexity.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113462"},"PeriodicalIF":7.2,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144480177","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Skeleton-based robust registration framework for corrupted 3D point clouds","authors":"Yongqiang Wang ,&nbsp;Weigang Li ,&nbsp;Wenping Liu ,&nbsp;Zhiqiang Tian ,&nbsp;Jinling Li","doi":"10.1016/j.asoc.2025.113461","DOIUrl":"10.1016/j.asoc.2025.113461","url":null,"abstract":"<div><div>Point cloud registration is fundamental in 3D vision applications, including autonomous driving, robotics, and medical imaging, where precise alignment of multiple point clouds is essential for accurate environment reconstruction. However, real-world point clouds are often affected by sensor limitations, environmental noise, and preprocessing errors, making registration challenging due to density distortions, noise contamination, and geometric deformations. Existing registration methods rely on direct point matching or surface feature extraction, which are highly susceptible to these corruptions and lead to reduced alignment accuracy. To address these challenges, a Skeleton-based Robust Registration Framework (SRRF) is presented, which introduces a corruption-resilient skeletal representation to improve registration robustness and accuracy. The framework integrates skeletal structures into the registration process and combines the transformations obtained from both the corrupted point cloud alignment and its skeleton alignment to achieve optimal registration. In addition, a distribution distance loss function is designed to enforce the consistency between the source and target skeletons, which significantly improves the registration performance. This framework ensures that the alignment considers both the original local geometric features and the global stability of the skeleton structure, resulting in robust and accurate registration results. Experimental evaluations on diverse corrupted datasets demonstrate that SRRF consistently outperforms state-of-the-art registration methods across various corruption scenarios, including density distortions, noise contamination, and geometric deformations. The results confirm the robustness of SRRF in handling corrupted point clouds, making it a potential approach for 3D perception tasks in real-world scenarios.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"181 ","pages":"Article 113461"},"PeriodicalIF":7.2,"publicationDate":"2025-06-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144491914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}