计算机科学最新文献

{"title":"Extreme Learning Machine Models for Classifying the LED Source in a 2D Visible Light Positioning Database","authors":"David Zabala-Blanco,&nbsp;Cesar A. Azurdia-Meza,&nbsp;Benjamín Lobos Soto,&nbsp;Ismael Soto,&nbsp;Pablo Palacios Játiva,&nbsp;Roberto Ahumada-García,&nbsp;Muhammad Ijaz","doi":"10.1049/ote2.70004","DOIUrl":"https://doi.org/10.1049/ote2.70004","url":null,"abstract":"<p>In recent years, there has been a surge in interest in indoor positioning systems that use visible light communication (VLC) technology combined with light-emitting diodes (LEDs). These systems have gained attention because of their ability to offer high bandwidth, precise localisation, and potential for wireless communication to extend into the visible light spectrum in the future, making VLC a notable candidate. Furthermore, the visible light spectrum proves advantageous in the industrial internet of things setting, as it does not offer electromagnetic interference as in radio frequency (RF) spectrum. This paper analyses a database made up of approximately 356 image samples obtained from a CMOS sensor. The database encompasses eight distinct classes, each demonstrating frequency (bit rate) variations ranging from 1 to 4.5 kHz in 500 Hz increments. The aim is to implement this database for classification applications as a first stage with several neural networks based on extreme learning machines (ELM) in various forms: (1) standard ELM, (2) regularised ELM, (3) weighted ELM in two configurations, and (4) multilayer ELM with 2 and 3 hidden layers. The findings of this study reveal that standard ELM is particularly promising, achieving more than 99% in accuracy and G-mean, while maintaining low computational complexity (measured in tenths of seconds) when compared to convolutional neural networks and multilayer perceptrons, which offer superior performance, however at the cost of significant computational demands.</p>","PeriodicalId":13408,"journal":{"name":"Iet Optoelectronics","volume":"19 1","pages":""},"PeriodicalIF":2.3,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/ote2.70004","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143877819","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CURRENT EVENTS","authors":"","doi":"10.1016/S0893-6080(25)00412-5","DOIUrl":"10.1016/S0893-6080(25)00412-5","url":null,"abstract":"","PeriodicalId":49763,"journal":{"name":"Neural Networks","volume":"187 ","pages":"Article 107533"},"PeriodicalIF":6.0,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143874191","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 Commit Classification Framework Incorporated With Prompt Tuning and External Knowledge","authors":"Jiajun Tong,&nbsp;Xiaobin Rui","doi":"10.1049/sfw2/5566134","DOIUrl":"https://doi.org/10.1049/sfw2/5566134","url":null,"abstract":"<div>\u0000 <p>Commit classification is an important task in software maintenance, since it helps software developers classify code changes into different types according to their nature and purpose. This allows them to better understand how their development efforts are progressing, identify areas where they need improvement, and make informed decisions about when and how to release new versions of their software. However, existing methods are all discriminative models, usually with complex architectures that require additional output layers to produce class label probabilities, making them task-specific and unable to learn features across different tasks. Moreover, they require a large amount of labeled data for fine tuning, and it is difficult to learn effective classification boundaries in the case of limited labeled data. To solve the above problems, we propose a generative framework that incorporates prompt tuning for commit classification with external knowledge (IPCK), which simplifies the model structure and learns features across different tasks, only based on the commit message information as the input. First, we proposed a generative framework based on T5 (text-to-text transfer transformer). This encoder–decoder construction method unifies different commit classification tasks into a text-to-text problem, simplifying the model’s structure by not requiring an extra output layer. Second, instead of fine tuning, we design a prompt tuning solution that can be adopted in few-shot scenarios with only limited samples. Furthermore, we incorporate external knowledge via an external knowledge graph to map the probabilities of words into the final labels in the speech machine step to improve performance in few-shot scenarios. Extensive experiments on two open available datasets demonstrate that our framework can solve the commit classification problem simply but effectively for both single-label binary classification and single-label multiclass classification purposes with 90% and 83% accuracy. Further, in the few-shot scenarios, our method improves the adaptability of the model without requiring a large number of training samples for fine tuning.</p>\u0000 </div>","PeriodicalId":50378,"journal":{"name":"IET Software","volume":"2025 1","pages":""},"PeriodicalIF":1.5,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/sfw2/5566134","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143875664","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A new hybrid data and model-centric predictive approach dedicated to industrial pipe maintenance","authors":"Ahmad Braydi ,&nbsp;Pascal Fossat ,&nbsp;Alessandro Casaburo ,&nbsp;Victor Pernet ,&nbsp;Cyril Zwick ,&nbsp;Mohsen Ardabilian ,&nbsp;Olivier Bareille","doi":"10.1016/j.engappai.2025.110821","DOIUrl":"10.1016/j.engappai.2025.110821","url":null,"abstract":"<div><div>Predictive Maintenance (PdM) for pipe clogging is a critical challenge in the industrial sector, particularly with the increasing adoption of Artificial Intelligence (AI) and the Internet of Things (IoT). Frequent clogging incidents, such as those faced by Orano/La Hague, lead to energy waste, operational inefficiencies, financial losses, and potential safety hazards, highlighting the critical need for effective maintenance solutions to protect both assets and personnel. This study proposes a novel hybrid approach that combines the strengths of data-centric and model-centric methodologies for Prognostic and Health Monitoring (PHM) of pipeline systems in constrained industrial environments. The approach utilizes passive acceleration measurements to predict clogging occurrences and quantify clogging severity under varying airflow rates. Experimental results indicate that the proposed method achieves up to 100% accuracy in clogging detection and robust performance across diverse operational conditions. This integrated methodology represents a significant step forward in predictive maintenance, offering scalable and adaptable solutions to enhance safety, operational efficiency, and cost-effectiveness in industrial settings.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"153 ","pages":"Article 110821"},"PeriodicalIF":7.5,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143874359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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":"Enhanced lightweight detection of small and tiny objects in high-resolution images using object tracking-based region of interest proposal","authors":"Aleksandra Kos ,&nbsp;Karol Majek ,&nbsp;Dominik Belter","doi":"10.1016/j.engappai.2025.110852","DOIUrl":"10.1016/j.engappai.2025.110852","url":null,"abstract":"<div><div>Detecting small objects in high-resolution images is challenging in practical applications. Downsampling the image makes small objects difficult or impossible to detect, while processing multiple low-resolution detection windows using a sliding-window approach is both time-consuming and impractical. To deal with this engineering problem, we present a novel window-based object detection system designed for detecting tiny and multi-scale objects in high-resolution images. Our approach introduces two Region of Interest (ROI) Modules that select full-resolution areas for the detector to focus on. We use a low-resolution current frame to estimate ROIs with a segmentation-based branch combined with past detection metadata to predict object locations with a tracking-based branch. By fusing outputs from these modules, we effectively recover regions with tiny objects overlooked by the estimation branch, and we significantly reduce the number of object detection runs compared to the sliding-window approach, maintaining the method’s speed. Moreover, we employ a fusion of downscaling and sliding-window techniques within large ROIs, complemented by our novel Overlapping Box Suppression (OBS) algorithm to reduce partial false-positive detections. We analyze our system and all its components on two challenging datasets — SeaDronesSee and DroneCrowd to show superior performance compared to state-of-the-art object detectors. Our approach enhances both the Artificial Intelligence (AI) and engineering domains by improving the quality and efficiency of tiny object detection, facilitating its integration into demanding real-time robotics applications. The inference code is available at <span><span>https://github.com/deepdrivepl/TinyROIFusion</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"153 ","pages":"Article 110852"},"PeriodicalIF":7.5,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143874360","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"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":"Climate-resilient water management: Leveraging IoT and AI for sustainable agriculture","authors":"Syed Muhammad Zaigham Abbas Naqvi ,&nbsp;Saddam Hussain ,&nbsp;Muhammad Awais ,&nbsp;Muhammad Naveed Tahir ,&nbsp;Shoaib Rashid Saleem ,&nbsp;Fuad A.M. Al-Yarimi ,&nbsp;Mirjalol Ashurov ,&nbsp;Oumaima Saidani ,&nbsp;M.Ijaz Khan ,&nbsp;Junfeng Wu ,&nbsp;Zhang Wei ,&nbsp;Jiandong Hu","doi":"10.1016/j.eij.2025.100691","DOIUrl":"10.1016/j.eij.2025.100691","url":null,"abstract":"<div><div>Climate change is the phenomenon of permanent change in the environmental conditions of an area. However, it is now affecting the earth by causing a permanent seasonal shift. This seasonal shift is not only decreasing the yields of crops by shortening their growth duration but also critically affecting the water availability for irrigation purposes. This article addresses the irrigation management strategies to mitigate the impacts of climate changes using advance techniques like internet of things (IoT). IoT is the setup of smart sensory devices which are interconnected using internet. They collect the data from field and analyze using artificial intelligence based algorithmic models. The irrigation management strategies using the artificial intelligence (AI) to mitigate the climate change impacts by reducing the wastage of essential resources in the environment has not been adopted by many developed countries. This article briefly explained the applications of AI in smart agriculture. Manuscript further describes the idea to protect the agricultural system from water scarcity and flooding by the efficient use of sensors, IoT and AI by automating the traditional agricultural practices. Different variable rate applications, smart irrigation methods like weather-based smart irrigation and moisture-based smart irrigation have been discussed in this review. Different countries have adapted different technologies of smart irrigation which can mitigate climate changes effectively and a case study with this respect is discussed. Moreover, implementations of integrated neural network models with the decision support system of irrigation management strategies to decide the supply of water in the field in real-time have been discussed in this review.</div></div>","PeriodicalId":56010,"journal":{"name":"Egyptian Informatics Journal","volume":"30 ","pages":"Article 100691"},"PeriodicalIF":5.0,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143876612","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"Optimized framework for strategic electric vehicle charging station placement and scheduling in distribution systems with renewable energy integration","authors":"Jajna Prasad Sahoo,&nbsp;S. Sivasubramani,&nbsp;Parimi Sai Syama Srikar","doi":"10.1016/j.swevo.2025.101943","DOIUrl":"10.1016/j.swevo.2025.101943","url":null,"abstract":"<div><div>Increased demand for electric vehicles (EVs) is faced with challenges by existing electrical grids. To address these challenges, in this study, a comprehensive framework is developed for the strategic placement of electric vehicle charging stations (EVCSs) in a distribution system and efficient charging and discharging schedules of EVs in the EVCSs. Optimal locations for EVCSs in a distribution system are identified with system inefficiencies and the inclusion of renewable energy sources (RESs), such as solar PV. EV scheduling is performed considering the power exchange between EVCSs and the grid with the integration of RESs in the distribution system. The framework is presented as an optimization problem and is addressed through the particle swarm optimization (PSO) approach. For comparison, the proposed model is also solved using the genetic algorithm (GA) and sine cosine algorithm (SCA). The IEEE 33 bus system is used as a test system to implement the suggested approach. The simulation outcomes show the effectiveness of the proposed model. The proposed PSO-based approach demonstrates significant improvements, reducing power losses by 10.29% for optimal EVCS placement compared to random placement, while also achieving cost reductions of 25.42% and 32% compared to SCA and GA, respectively, through optimized EVCS placement and scheduling. Validation through real-time implementation is performed using the OPAL-RT platform. The experimental setup confirms the real-time feasibility of the suggested approach.</div></div>","PeriodicalId":48682,"journal":{"name":"Swarm and Evolutionary Computation","volume":"95 ","pages":"Article 101943"},"PeriodicalIF":8.2,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873585","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":"Large vision-language models enabled novel objects 6D pose estimation for human-robot collaboration","authors":"Wanqing Xia,&nbsp;Hao Zheng,&nbsp;Weiliang Xu,&nbsp;Xun Xu","doi":"10.1016/j.rcim.2025.103030","DOIUrl":"10.1016/j.rcim.2025.103030","url":null,"abstract":"<div><div>Six-Degree-of-Freedom (6D) pose estimation is essential for robotic manipulation tasks, especially in human-robot collaboration environments. Recently, 6D pose estimation has been extended from seen objects to novel objects due to the frequent encounters with unfamiliar items in real-life scenarios. This paper presents a three-stage pipeline for 6D pose estimation of previously unseen objects, leveraging the capabilities of large vision-language models. Our approach consists of vision-language model-based object detection and segmentation, mask selection with pose hypothesis generated from CAD models, and refinement and scoring of pose candidates. We evaluate our method on the YCB-Video dataset, achieving a state-of-the-art Average Recall (AR) score of 75.8 with RGB-D images, demonstrating its effectiveness in accurately estimating 6D poses for a diverse range of objects. The effectiveness of each operation stage is investigated in the ablation study. To validate the practical applicability of our approach, we conduct case studies on a real-world robotic platform, focusing on object pick-up tasks by integrating our 6D pose estimation pipeline with human intention prediction and task analysis algorithms. Results show that the proposed method can effectively handle novel objects in our test environments, as demonstrated through the YCB dataset evaluation and case studies. Our work contributes to the field of human-robot collaboration by introducing a flexible, generalizable approach to 6D pose estimation, enabling robots to adapt to new objects without requiring extensive retraining—a vital capability for advancing human-robot collaboration in dynamic environments. More information can be found in the project GitHub page: <span><span>https://github.com/WanqingXia/HRC_DetAnyPose</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":21452,"journal":{"name":"Robotics and Computer-integrated Manufacturing","volume":"95 ","pages":"Article 103030"},"PeriodicalIF":9.1,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143874159","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":"Optimizing wind energy integration with fuzzy PID DSTACOM and hybrid MPPT for reduced harmonic distortion","authors":"Shinagam Rajshekar,&nbsp;Lalit Chandra Saikia,&nbsp;Lavanya Nandyala","doi":"10.1016/j.compeleceng.2025.110391","DOIUrl":"10.1016/j.compeleceng.2025.110391","url":null,"abstract":"<div><div>Wind Energy Conversion Systems (WECS) play a crucial role in renewable energy generation but face challenges such as fluctuating wind speeds and grid instability, leading to power quality issues and voltage fluctuations. This study proposes an advanced integration of a wind-connected grid with a Battery Energy Storage System (BESS), optimized through a fuzzy PID-controlled DSTATCOM and a hybrid Maximum Power Point Tracking (MPPT) algorithm combining Perturb and Observe (P&amp;O) with Incremental Conductance (InCon). The fuzzy PID-controlled DSTATCOM provides dynamic reactive power compensation, stabilizes voltage, and mitigates harmonic distortions, while the BESS ensures efficient energy storage and dispatch under varying wind conditions. Simulation results demonstrate a significant reduction in Total Harmonic Distortion (THD), with the DFIG current THD peaking at 10 % before stabilization, while grid and load currents were reduced to below 0.15 % and 0.02 %, respectively. The findings confirm that the proposed system enhances power extraction, minimizes harmonic distortions, and improves grid stability. These results suggest that the developed methodology provides a robust and efficient solution for integrating wind energy into modern smart grids.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"124 ","pages":"Article 110391"},"PeriodicalIF":4.0,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143873706","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"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":"Horizontal-to-tilted conversion of solar radiation data using machine learning algorithms","authors":"Ali Naci Celik ,&nbsp;Bahadır Sarman ,&nbsp;Kemal Polat","doi":"10.1016/j.engappai.2025.110951","DOIUrl":"10.1016/j.engappai.2025.110951","url":null,"abstract":"<div><div>Solar radiation is the main input of system design algorithms in solar energy engineering. Solar radiation is usually measured on horizontal surfaces. However, in majority of solar energy applications such as photovoltaics, surfaces are either fixed at certain angles or continuously track the sun for maximizing energy input. Therefore, converting solar radiation data from horizontal to tilted surfaces is essential. Conventionally, conversion of solar radiation from horizontal to tilted is carried out using analytical methods. As with many other disciplines in science and technology, machine learning has recently been successfully applied also to solar radiation modelling to solve various problems such as in-advance forecasting of solar radiation. In the present article, solar radiation collected on horizontal surface is converted to tilted surface by machine learning algorithms and compared to solar radiation measured at a tilted surface. Eight different machine learning algorithms have been presently used for the conversion of solar radiation data. Accuracy of the models has been assessed based on a total of seven statistical metrics commonly used in literature. Overall, extra trees algorithm led to the best results as indicated by the statistical metrics used, for example, the mean absolute error of 7.3219 and coefficient of determination 0.9964. Based on the results presently obtained, it is demonstrated that machine learning led to an improved prediction when compared to the analytical models. The present research highlights the crucial significance of such advanced techniques, emphasizing their potential to drive a paradigm shift in solar energy engineering.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"153 ","pages":"Article 110951"},"PeriodicalIF":7.5,"publicationDate":"2025-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143874356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}