Computational Intelligence最新文献

{"title":"Dynamic Object Detection and Tracking System on Unmanned Aerial Vehicles for Surveillance Applications Using RegionViT-Based Adaptive Multi-Scale YOLOv8","authors":"Venkateswara Raju Yallamraju,&nbsp;Selvaganesan Jana","doi":"10.1111/coin.70101","DOIUrl":"https://doi.org/10.1111/coin.70101","url":null,"abstract":"<div>\u0000 \u0000 <p>In general, the object detection mechanism recognizes target objects in the image frames, and the tracking mechanism helps to capture the movement of target objects in diverse frames. Recent developments in Artificial Intelligence (AI) have enabled us to build computers, robots, and automated tools that are mostly designed for performing tasks and generating decisions without human assistance. The drones called Unmanned Aerial Vehicles (UAVs) are employed for many kinds of objectives, including parcel shipping, rescue operations, recognizing objects, and monitoring. Object detection and tracking systems are crucial for UAVs because they help to optimally capture moving objects, areas, and threats for enhancing security, surveillance, and awareness at an earlier stage. Also, they help to ultimately predict the category and location of UAVs from the video frames. Many researchers have developed diverse object detection and tracking methods on UAVs; yet, it is complex for continuously monitoring small objects in the gathered data, and it is affected by noise and blurriness due to the motion of UAVs. One of the greatest challenging duties for UAVs is object recognition and tracking since it needs precise, swift, and cost-effective object detection and tracking. Pre-trained networks are required for the detection of objects based on deep learning. Mismatches between the pre-trained and the target domain network areas create problems in object detection. With the aim of resolving these issues, a deep learning-assisted UAV control mechanism is developed in this research work by performing detection and tracking of objects. The developed model is helpful in improving rescue operations in disaster areas and security surveillance. At first, the input videos are accumulated from benchmark sources. From the videos, the resolution of the images is studied for an accurate object detection procedure. Next, the detection and tracking of the object are done via the developed Region Vision Transformer-based Adaptive Multi-scale You Only Look Once v8 (RV-AMYOLOv8). The parameters fromYOLOv8 are optimized using the Fitness-based Random Variable for Elk Herd Optimizer (FRVEHO) for enhancing the performance. The quantitative outcomes of the implemented approach help to analyze the suggested network's performance with conventional techniques. Here, the developed method attains 96% accuracy and 95% of precision measure to demonstrate its better performance than the existing methods. This object detection is helpful for analyzing the surrounding obstacles while controlling the UAVs.</p>\u0000 </div>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144598389","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Does Linguistic Relativity Hypothesis Apply on ChatGPT Responses? Yes, It Does","authors":"Partha Pratim Ray","doi":"10.1111/coin.70103","DOIUrl":"https://doi.org/10.1111/coin.70103","url":null,"abstract":"&lt;div&gt;\u0000 \u0000 &lt;p&gt;We present the first comprehensive, end-to-end quantitative evaluation of the linguistic relativity hypothesis in AI-generated text, using ChatGPT-4o mini to generate responses to 10 culturally salient prompts across 13 typologically diverse languages. Semantic shifts were quantified using pairwise cosine similarity scores computed from multilingual MiniLM sentence embeddings. A one-way analysis of variance (ANOVA) reveals statistically significant variation in semantic alignment across language pairs, with &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;F&lt;/mi&gt;\u0000 &lt;mo&gt;(&lt;/mo&gt;\u0000 &lt;mn&gt;77&lt;/mn&gt;\u0000 &lt;mo&gt;,&lt;/mo&gt;\u0000 &lt;mn&gt;702&lt;/mn&gt;\u0000 &lt;mo&gt;)&lt;/mo&gt;\u0000 &lt;mo&gt;=&lt;/mo&gt;\u0000 &lt;mn&gt;2&lt;/mn&gt;\u0000 &lt;mo&gt;.&lt;/mo&gt;\u0000 &lt;mn&gt;153&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;annotation&gt;$$ Fleft(77,702right)=2.153 $$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;, &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;p&lt;/mi&gt;\u0000 &lt;mo&gt;=&lt;/mo&gt;\u0000 &lt;mn&gt;2&lt;/mn&gt;\u0000 &lt;mo&gt;.&lt;/mo&gt;\u0000 &lt;mn&gt;29&lt;/mn&gt;\u0000 &lt;mo&gt;×&lt;/mo&gt;\u0000 &lt;mn&gt;1&lt;/mn&gt;\u0000 &lt;msup&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mn&gt;0&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mo&gt;−&lt;/mo&gt;\u0000 &lt;mn&gt;7&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;/msup&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;annotation&gt;$$ p=2.29times 1{0}^{-7} $$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;, and effect size &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;msup&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;η&lt;/mi&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mn&gt;2&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;/msup&gt;\u0000 &lt;mo&gt;=&lt;/mo&gt;\u0000 &lt;mn&gt;0&lt;/mn&gt;\u0000 &lt;mo&gt;.&lt;/mo&gt;\u0000 &lt;mn&gt;191&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;annotation&gt;$$ {eta}^2=0.191 $$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;. These results are further supported by a non-parametric Kruskal–Wallis test yielding &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;H&lt;/mi&gt;\u0000 &lt;mo&gt;=&lt;/mo&gt;\u0000 &lt;mn&gt;176&lt;/mn&gt;\u0000 &lt;mo&gt;.&lt;/mo&gt;\u0000 &lt;mn&gt;208&lt;/mn&gt;\u0000 &lt;/mrow&gt;\u0000 &lt;annotation&gt;$$ H=176.208 $$&lt;/annotation&gt;\u0000 &lt;/semantics&gt;&lt;/math&gt;, &lt;span&gt;&lt;/span&gt;&lt;math&gt;\u0000 &lt;semantics&gt;\u0000 &lt;mrow&gt;\u0000 &lt;mi&gt;p&lt;/mi&gt;\u0000 &lt;mo&gt;=&lt;/mo&gt;\u0000 &lt;mn&gt;9&lt;/mn&gt;\u0000 &lt;mo&gt;.&lt;/mo&gt;\u0000 &lt;mn&gt;59&lt;/mn&gt;\u0000 &lt;mo&gt;×&lt;/mo&gt;\u0000 &lt;mn&gt;1&lt;/mn&gt;\u0000 &lt;msup&gt;\u0000 &lt;mrow&gt;\u0000 ","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-07-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144598683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"SparseMult: A Sparse Tensor Decomposition Model for Knowledge Graph Link Prediction","authors":"Zhiwen Xie,&nbsp;Runjie Zhu,&nbsp;Meng Zhang,&nbsp;Jin Liu","doi":"10.1111/coin.70097","DOIUrl":"https://doi.org/10.1111/coin.70097","url":null,"abstract":"<div>\u0000 \u0000 <p>Knowledge graphs (KGs) have shown great power in many downstream natural language processing (NLP) tasks, such as recommendation system and question answering. Despite the large amount of knowledge facts in KGs, KGs still suffer from an issue of incompleteness, namely, lots of relations between entities are missing. Link prediction, also known as knowledge graph completion (KGC), aims to predict missing relations between entities. The models based on tensor decomposition, such as Rescal and DistMult, are promising to solve the link prediction task. However, previous Rescal model lacks the ability to scale to large KGs due to the large amount of parameters. DistMult simplifies Rescal by using diagonal matrices to represent relations, while it suffers from the limitation of dealing with antisymmetric relations. To address these problems, in this paper, we propose a SparseMult model, which is a novel tensor decomposition model based on sparse relation matrix. Specifically, we view KGs as 3D tensors and decompose them as entity vectors and relation matrices. To reduce the number of parameters in relation matrices, we represent each relation matrix as a sparse block diagonal matrix. Thus, the complexity of relation matrices grow linearly with the embedding size, making it able to scale up to large KGs. Moreover, we analyze the ability of modeling different relation patterns and show that our SparseMult is capable to model symmetry, antisymmetry, and inversion relations. We conduct extensive experiments on three widely used benchmark datasets FB15k-237, WN18RR, and CCKS2021 KGs. Experimental results demonstrate that our SparseMult model outperforms most of the state-of-the-art methods.</p>\u0000 </div>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144582373","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Generalizing and Classifying From Few Samples: A Comprehension of Approaches to Few-Shot Visual Learning","authors":"Nadeem Yousuf Khanday,&nbsp;Shabir Ahmad Sofi","doi":"10.1111/coin.70098","DOIUrl":"https://doi.org/10.1111/coin.70098","url":null,"abstract":"<div>\u0000 \u0000 <p>Unlike traditional machine learning techniques, few-shot learning (FSL) represents a paradigm aimed at acquiring new tasks from just a handful of labeled examples. The challenge in FSL lies in its requirement for models to generalize effectively from a small dataset to previously unseen examples. Various approaches have been developed for FSL, encompassing techniques such as metric learning, meta-learning, and hybrid methods, among others. These approaches have found success in numerous computer vision tasks, including image and video classification, object detection, object segmentation, robotics, natural language processing, and various real-world applications such as medical diagnosis and self-driving cars. This comprehensive survey offers an in-depth exploration of recent advancements and the current state-of-the-art in FSL. The study presents a thorough examination of different FSL approaches, categorizing them primarily into meta-learning and non-meta-learning methods. It also delves into benchmark datasets for FSL, highlights existing research challenges, and explores the diverse applications of FSL. Furthermore, the survey identifies and discusses open research challenges within the field of FSL.</p>\u0000 </div>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144558122","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Zeroing Neural Network for Real-Time Operational Research and Computational Intelligence: An Ordinary Differential Equation Based Approach","authors":"Xinwei Cao,&nbsp;Penglei Li,&nbsp;Yufei Wang,&nbsp;Cheng Hua,&nbsp;Ameer Tamoor Khan","doi":"10.1111/coin.70099","DOIUrl":"https://doi.org/10.1111/coin.70099","url":null,"abstract":"<div>\u0000 \u0000 <p>The zeroing neural network (ZNN), a canonical recurrent neural network, was developed in previous studies to address time-varying problem-solving scenarios. Numerous practical applications involve time-varying linear equations and inequality systems that demand real-time solutions. This article proposes a ZNN model specifically designed to solve such time-varying linear systems. Innovatively, it incorporates a new non-negative slack variable that transforms complex time-varying inequality systems into more easily solvable time-varying equation systems. By using an exponential decay formula and establishing an indefinite error function, the ZNN model is built. The suggested ZNN model's convergence properties are validated by theoretical research. Results from comparative simulations further support the superiority and effectiveness of the ZNN model in resolving inequality systems and time-varying linear equations.</p>\u0000 </div>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144550934","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"AutoMathKG: The Automated Mathematical Knowledge Graph Based on LLM and Vector Database","authors":"Rong Bian,&nbsp;Yu Geng,&nbsp;Zijian Yang,&nbsp;Bing Cheng","doi":"10.1111/coin.70096","DOIUrl":"https://doi.org/10.1111/coin.70096","url":null,"abstract":"<div>\u0000 \u0000 <p>A mathematical knowledge graph (KG) presents knowledge within mathematics in a structured manner. Constructing a math KG using natural language is an essential but challenging task. Existing methods have two major limitations: Incomplete knowledge due to limited corpora and a lack of fully automated integration of diverse sources. This paper proposes AutoMathKG, a high-quality, wide-coverage, and multi-dimensional math KG capable of automatic updates. AutoMathKG regards mathematics as a vast directed graph composed of Definition, Theorem, and Problem entities, with their reference relationships as edges. It integrates knowledge from ProofWiki, textbooks, arXiv papers, and TheoremQA, enhanced through large language models (LLMs) for data augmentation. To search for similar entities, MathVD, a vector database, is built through two designed embedding strategies. To automatically update, two mechanisms are proposed. For knowledge completion, Math LLM is developed to interact with AutoMathKG, providing missing proofs or solutions. For knowledge fusion, MathVD is used to retrieve similar entities, and LLM is used to determine whether to merge with a candidate or add a new entity. Extensive experiments demonstrate the advanced performance of the AutoMathKG system, including superior reachability query results in MathVD compared to five baselines and robust mathematical reasoning capability in Math LLM.</p>\u0000 </div>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-07-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144536844","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Correction to “Short-Term Wind Speed Prediction Model Based on Hybrid Decomposition Method and Deep Learning”","authors":"","doi":"10.1111/coin.70100","DOIUrl":"https://doi.org/10.1111/coin.70100","url":null,"abstract":"<p>X. Yuan, F. Hu, and Z. Zhu, “Short-Term Wind Speed Prediction Model Based on Hybrid Decomposition Method and Deep Learning,” <i>Computational Intelligence</i> 41 (2025): e70078, https://doi.org/10.1111/coin.70078.</p><p>In the paper by Yuan and Zhu et al. (2025), the affiliation address “School of Electronic Engineering, Hunan College of Information, Xiangtan, China” of Xueqiong Yuan was incorrect. This should have read: “School of Electronic Engineering, Hunan College of Information, Changsha, China” (Xiangtan has been changed to Changsha).</p>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-06-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1111/coin.70100","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144514979","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":"Spprnet: A Robust CNN for Library Material Recognition via Spatial Pyramid Pooling and Heterogeneous Convolution","authors":"Li FeiFei,&nbsp;Meng Qi,&nbsp;Hong Bo,&nbsp;Zhang Lixiang,&nbsp;Ji Wen","doi":"10.1111/coin.70094","DOIUrl":"https://doi.org/10.1111/coin.70094","url":null,"abstract":"<div>\u0000 \u0000 <p>In library environments, the diverse scales of materials and batch instability caused by inconsistent scanning conditions pose challenges for image recognition tasks. Traditional ResNet architectures, due to their fixed input size constraints, may reduce their recognition accuracy for images of arbitrary sizes. In this study, we introduce a novel heterogeneous convolution strategy, adjust batch normalization operations, and incorporate a spatial pyramid pooling module based on the ResNet18 network to eliminate these limitations. This new architecture, termed SPPRNet, supports flexible processing of arbitrary-sized inputs and combines multi-scale convolution kernels (3 <span></span><math>\u0000 <semantics>\u0000 <mrow>\u0000 <mo>×</mo>\u0000 </mrow>\u0000 <annotation>$$ times $$</annotation>\u0000 </semantics></math> 3, 5 <span></span><math>\u0000 <semantics>\u0000 <mrow>\u0000 <mo>×</mo>\u0000 </mrow>\u0000 <annotation>$$ times $$</annotation>\u0000 </semantics></math> 5, 7 <span></span><math>\u0000 <semantics>\u0000 <mrow>\u0000 <mo>×</mo>\u0000 </mrow>\u0000 <annotation>$$ times $$</annotation>\u0000 </semantics></math> 7) to simultaneously capture fine-grained features and global contextual patterns. Quantitative results on general datasets demonstrate that our method achieves a 25.47% Top-1 error rate on ImageNet (compared to 30.55% for ResNet18) and attains 92.95% mAP on the Caltech-101 dataset for object detection tasks, outperforming mainstream models such as VGG-16 and MobileNet. The robust performance of our method in image tasks can be extended to existing approaches to further improve the quality of image recognition in library scenarios.</p>\u0000 </div>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144482068","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Frequency-Driven Diffusion: A Hierarchical Attention Weighting Framework for Underwater Image Restoration","authors":"Longxiang Deng,&nbsp;Laibin Chang,&nbsp;Wei Liu","doi":"10.1111/coin.70095","DOIUrl":"https://doi.org/10.1111/coin.70095","url":null,"abstract":"<div>\u0000 \u0000 <p>Underwater images often suffer from visual degradation, affecting downstream tasks. While recent underwater image enhancement (UIE) techniques have made some advances benefiting from deep neural networks, challenges remain in restoring fine details and achieving computational efficiency. Inspired by the success of diffusion models in image generation, we propose the Underwater Laplacian-Guided Diffusion Model (ULDM), which enhances image features layer-by-layer based on the hierarchical structure of the Laplacian pyramid transform to achieve both high-quality and efficient UIE. The Laplacian pyramid decomposes the degraded image into high- and low-frequency components, enabling the model to denoise the low-frequency spectrum and address global image degradation, thereby reducing computational overhead. To efficiently enhance high-frequency details, we introduce the Hierarchical Attention Weighted Module (HAWM) that leverages the strong pixel correlations in high-frequency sub-images at different levels, adjusting them layer-by-layer to better capture fine details. These high-frequency sub-images exhibit strong pixel correlation and consistent texture features across different layers, and their hierarchical pattern ensures effective detail restoration. Extensive experiments demonstrate that ULDM outperforms state-of-the-art methods in both quantitative and qualitative evaluations.</p>\u0000 </div>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144482069","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"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":"Privacy-Preserving and Efficient Pneumonia Diseases Detection System Based on Federal Intelligent Edges","authors":"Haoda Wang,&nbsp;Chen Qiu,&nbsp;Guowei Liu,&nbsp;Chunhua Su","doi":"10.1111/coin.70076","DOIUrl":"https://doi.org/10.1111/coin.70076","url":null,"abstract":"<div>\u0000 \u0000 <p>As pneumonia cases continue to rise worldwide, rapid diagnostic capabilities are essential for effective treatment. However, traditional medical systems often lack efficiency and coordinated management. In response, we propose an AI-driven biomedical diagnosis platform for real-time detection and swift intervention. Leveraging privacy-preserving deep learning on the edge, users can promptly obtain automated diagnoses by uploading chest CT images. To further enhance accuracy, we employ a federated learning (FL) framework that ensures scalable training in an industrial IoT setting while protecting patient data. Our global FL model achieves around 96.25% accuracy on a validation dataset, outperforming individual clients by 3.42%. By eliminating the need for sharing raw data, patient privacy is preserved, and the system offers improved flexibility and scalability for medical diagnosis.</p>\u0000 </div>","PeriodicalId":55228,"journal":{"name":"Computational Intelligence","volume":"41 4","pages":""},"PeriodicalIF":1.8,"publicationDate":"2025-06-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144472967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}