Expert Systems with Applications最新文献

{"title":"Text-prompted generative data augmentation and semi-supervised learning for indoor defect detection using quadruped robots","authors":"Xiang Ji ,&nbsp;Junjie Chen ,&nbsp;Yonglin Fu ,&nbsp;Isabelle Chan ,&nbsp;Zhen Dong","doi":"10.1016/j.eswa.2026.131545","DOIUrl":"10.1016/j.eswa.2026.131545","url":null,"abstract":"<div><div>Timely defect detection is the key to formulating targeted maintenance plans to extend a facility’s lifespan. While tremendous efforts have been made in deploying robots (e.g., drones) for outdoor defect detection, little attention has been paid to defects taking place indoors. Indoor defect detection (IDD) has distinctive characteristics concerning (a) the complex environment (narrow passages, staircases, etc.) that challenges inspection data collection, and (b) the drastic image feature variation caused by uneven illumination and view point changes, which renders methods viable for outdoor detection less useful. This research takes on the challenges and proposes an automated IDD approach. To navigate challenging indoor environments (e.g., staircases), a quadruped robot platform is proposed for inspection image collection. To address the scarcity of indoor data, a novel algorithmic framework for IDD is formulated that integrates large generative models for data augmentation and semi-supervised learning to train on the generated unlabeled data. It is found that the proposed approach can effectively inspect challenging indoor space for defect detection by leveraging the unique locomotion capability of legged robots. Despite the lack of training data, the framework resulted in a performance gain of 5.03% for the model. Future research is suggested to explore autonomous navigation of the robots and three dimensional modeling of the detected defects.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131545"},"PeriodicalIF":7.5,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192849","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 deep learning-based multi-modal approach to robust place recognition in challenging orchard environments","authors":"Dilshan Ranasinghe,&nbsp;Chao Chen","doi":"10.1016/j.eswa.2026.131543","DOIUrl":"10.1016/j.eswa.2026.131543","url":null,"abstract":"<div><div>Place recognition addresses the problem of identifying the current location on revisit using the acquired knowledge of all the previously explored locations. The abstract representations of the locations can be generated as they are being explored using processed single or multi-sensory inputs. Later, the current abstraction of the environment can be compared with previously collected information to detect revisits. The loop closure component of Simultaneous Localization And Mapping (SLAM) uses these techniques to identify revisits, which allow the SLAM system to mitigate the associated long-term drift. RGB images, LiDAR point clouds, and LiDAR intensities are the three most popular sensory inputs used in place recognition literature. However, they each have certain disadvantages when used as a single-modal input. Additionally, in recent years, Deep Neural Network (DNN) based methods have emerged increasingly in the literature that addresses the place recognition problem. Therefore, In this work, we introduce a novel multi-modal method that takes advantage of the rich complementary information provided by the above three modalities along with a DNN for place recognition. This work was evaluated on multiple publicly available datasets as well as on a highly repetitive orchard dataset collected by our team. The results demonstrate the ability of this method to be used even in highly challenging environments such as orchards.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131543"},"PeriodicalIF":7.5,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192700","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":"High-performance clustering and routing for industrial wireless sensor networks: An innovative heuristic optimization approach","authors":"Jing Xiao ,&nbsp;Mengfei Wang ,&nbsp;Tao Luo ,&nbsp;Zhigang Li","doi":"10.1016/j.eswa.2026.131408","DOIUrl":"10.1016/j.eswa.2026.131408","url":null,"abstract":"<div><div>Industrial wireless sensor networks (IWSNs) play a critical role in enabling reliable data acquisition in industrial environments, where routing scheme directly affects network lifetime and communication reliability. This paper presents a novel clustering routing model that integrates node residual energy, clustering distance, and link quality to better reflect practical network conditions. Based on this model, a high-performance cluster routing protocol leveraging a quantum chaotic immune clone algorithm (HCR-QCICA) is proposed. The protocol employs a new chaotic initialization strategy to avoid optima and a novel quantum optimization strategy to enhance global search capability. Additionally, a type-differentiated inter-cluster multi-hop routing method is developed to address diverse data transmission requirements. Experimental results in simulated industrial scenarios show that HCR-QCICA reduces delay and packet loss rate by at least 4.08% and 9.16%, respectively, while increasing network lifetime and throughput by over 14.07% and 24.08%, compared to LEACH, TEEN, CHEABC-QCRP, MOGWO, and GAPSO-H protocols. These findings demonstrate the effectiveness of the proposed approach in improving IWSNs performance.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131408"},"PeriodicalIF":7.5,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192845","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":"Quantifying expert speech: A comprehensive analysis of instructional discourses","authors":"Mariano Albaladejo-González ,&nbsp;Manuel J. Gomez ,&nbsp;Óscar Cánovas ,&nbsp;Félix Gómez Mármol ,&nbsp;José A. Ruipérez-Valiente","doi":"10.1016/j.eswa.2026.131515","DOIUrl":"10.1016/j.eswa.2026.131515","url":null,"abstract":"<div><div>Oral communication is a crucial skill in modern society. Nevertheless, it requires sustained practice and constructive feedback. Consequently, several studies have explored the development of oral communication trainers powered by Artificial Intelligence (AI). However, what characterizes expert speech remains unclear, especially given the need to adapt speech to contextual factors. In instructional environments, the speaker’s communication proficiency is a key determinant of audience learning outcomes. For this reason, we have analyzed 1250 speeches from five types of instructional discourses: in-person college classes (<em>Lectures</em>), online learning lessons (<em>Online Courses</em>), instructional animations (<em>Animated Lessons</em>), supplementary materials for school and high school (<em>Supplementary Lessons</em>), and public presentations (<em>Public Talks</em>). We extracted 16 speech metrics, including six additional multiple-participant metrics for <em>Lectures</em>. We obtained 250 videos of each discourse type, ensuring a minimum length of five minutes. Our analysis revealed expert values for each speech metric and showed how speech metrics vary across discourse types. We also developed an AI speech classifier that achieved an F1 score of 0.78. The model struggled to identify <em>Online Courses</em>, which is consistent with the Uniform Manifold Approximation and Projection analysis, showing that <em>Online Courses</em> are closely interjected with the speech of other instructional discourses. Furthermore, we identified distinct speech profiles in <em>Lectures, Public Talks</em>, and <em>Online Courses</em>, highlighting variations in speaking styles. This research provides valuable insights into expert speech in instructional discourses by offering reference values that can help speakers refine their delivery and support researchers in developing more effective speech training systems.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131515"},"PeriodicalIF":7.5,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192847","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":"Integrating diversity-integrated weighted ranking in metaheuristic algorithms for medical applications","authors":"Qinghong Hou ,&nbsp;Qike Shao ,&nbsp;Ali Asghar Heidari ,&nbsp;Lei Liu ,&nbsp;Huiling Chen ,&nbsp;Guoxi Liang","doi":"10.1016/j.eswa.2026.131301","DOIUrl":"10.1016/j.eswa.2026.131301","url":null,"abstract":"<div><div>Metaheuristic algorithms have become essential for addressing high-dimensional medical optimization problems, yet many still suffer from premature convergence and limited precision. To overcome these issues, we propose a diversity-integrated weighted ranking strategy that combines fitness and population diversity to construct a novel ranking matrix, reshaping convergence dynamics and achieving a better balance between exploration and exploitation. On the IEEE CEC 2017 benchmark suite, this strategy achieves significant improvements on 29 of 30 functions. Building on this foundation, we integrate the mechanism into the Artemisinin Optimization Algorithm and develop WRAO, which is further applied to two key medical tasks: feature selection and medical image segmentation. In feature selection, WRAO effectively identifies compact and discriminative medical features, enhancing disease classification accuracy and interpretability. In image segmentation, it optimizes boundary localization and structural consistency across varying imaging conditions, facilitating more precise pathological analysis. These applications demonstrate that WRAO not only advances metaheuristic optimization but also provides a robust and adaptable computational tool for medical data analysis and decision support, capable of meeting the precision and reliability demands of complex clinical environments.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"311 ","pages":"Article 131301"},"PeriodicalIF":7.5,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190987","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":"Compressive sensing image restoration with deep prior guided group sparse representation","authors":"Zhulin Ji ,&nbsp;Shenghai Liao ,&nbsp;Ruyi Han ,&nbsp;Shujun Fu","doi":"10.1016/j.eswa.2026.131465","DOIUrl":"10.1016/j.eswa.2026.131465","url":null,"abstract":"<div><div>Compressive sensing (CS) enables accurate reconstruction of images from significantly fewer measurements than required by the Nyquist-Shannon sampling theorem, relying critically on effective image priors to regularize the ill-posed inverse problem. Conventional patch-based sparse representation utilize fixed dictionaries that are learned off-the-shelf using the K-SVD algorithm. However, patch-based sparse representation ignores the relationship among patches, and the learned dictionaries can not capture the global image statistics, which will lead to suboptimal reconstruction performance. In this paper, we exploit group sparse representation (GSR) for image compressive sensing reconstruction. By clustering non-local image patches into group and regarding each group as a unit, group sparse representation simultaneously finding sparse codes for all patches within a group, leading to improved reconstruction fidelity and edge preservation. However, GSR relies solely on the undersampled image itself to construct dictionary that is not learnable, being increasingly unreliable at low compressive sensing rates where substantial loss of local image information occurs. To address this limitation, we propose a Deep Prior guided Group Sparse Representation (DPGSR) model for compressive image restoration, where a deep denoiser is responsible for capturing and learning both local and global image statistics by training on external data. The proposed DPGSR achieves improved global consistency, effectively reducing block artifacts while preserving sharper local details. Extensive experiments on image compressive sensing reconstruction and fast MRI demonstrate that the proposed method outperforms state-of-the-art approaches, particularly in preserving fine details and reducing over-smoothing artifacts.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131465"},"PeriodicalIF":7.5,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192781","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":"HHGDroid: Hybrid heterogeneous graph-based android malware detection via multi-evidence similarity fusion","authors":"Junwei Tang ,&nbsp;Xiaomei Tian ,&nbsp;Tao Peng ,&nbsp;Jianfeng Lu ,&nbsp;Haozhao Wang ,&nbsp;Ruixuan Li","doi":"10.1016/j.eswa.2026.131528","DOIUrl":"10.1016/j.eswa.2026.131528","url":null,"abstract":"<div><div>Currently, static analysis is insufficient to deal with Android malware that employs advanced evasion techniques such as code obfuscation and dynamic loading. Therefore, hybrid analysis that combines static structure and dynamic behavior has become the mainstream trend. However, existing hybrid analysis methods often adopt simple feature concatenation or shallow fusion mechanisms, which cannot effectively integrate heterogeneous static and dynamic features or capture the complex correlations between structure and behavior. To address this, we propose a hybrid heterogeneous graph-based Android malware detection method via multi-evidence similarity fusion, named HHGDroid. The function call graph generated by static analysis and the event graph obtained through dynamic analysis are connected through a comprehensive similarity of multiple evidences such as semantics, permissions, and time frequency, ultimately forming the hybrid heterogeneous graph with multiple heterogeneous nodes and edges. Our constructed hybrid heterogeneous graph is the first one that simultaneously possesses static and dynamic features. Finally, we improve Reliability-Calibrated Heterogeneous Graph Transformer (RCHGT) to learn the multiple relationships in the hybrid heterogeneous graph, which can automatically distinguish reliable and unreliable edges during the information propagation stage. We conduct experiments on real Android malware applications and achieved an F1-score of 97.87%, outperforming the state-of-the-art methods. Additionally, we verify our method on an unknown malware dataset and obtained an F1-score of 81.52%, which is superior to existing methods. HHGDroid is a novel and effective method for detecting Android malware.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131528"},"PeriodicalIF":7.5,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192853","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":"Surv-RWKV: Cross-modal receptance weighted key-value interaction with optimal transport feature alignment for survival analysis","authors":"Xiyang Kuang ,&nbsp;Bin Yang ,&nbsp;Bingo Wing-Kuen Ling ,&nbsp;Kok Lay Teo ,&nbsp;Xiaozhi Zhang","doi":"10.1016/j.eswa.2026.131506","DOIUrl":"10.1016/j.eswa.2026.131506","url":null,"abstract":"<div><div>Multimodal learning has played a pivotal role in survival prediction, particularly in integrating pathological images and genomic data for improving predictive performance. Pathological images provide macroscopic histological information about tumor morphology, while genomic data reveal molecular-level genetic characteristics. The integration of these two modalities enables a comprehensive characterization of tumor heterogeneity and disease progression mechanisms. Despite recent advances in multimodal integration that have significantly enhanced prognostic accuracy, challenges remain in effectively analyzing high-dimensional and heterogeneous whole-slide images and omics data. Current Transformer-based sequence modeling approaches suffer from limited computational efficiency when processing long feature sequences and capturing complex cross-modal interactions. To address these challenges, we propose an innovative cross-modal receptance weighted key-value (RWKV)-based framework, termed Surv-RWKV, for survival prediction. This framework integrates RWKV-based sequence modeling with advanced multimodal fusion strategies to enhance both predictive accuracy and model efficiency. Specifically, Surv-RWKV employs parallel RWKV-based encoders to model long-range dependencies in WSI tissue cluster patterns and genomic pathway activation profiles, achieving improved prognostic performance with optimized computational efficiency. Subsequently, a transport-based optimal cross-modal alignment module is introduced to establish semantic correspondences between histopathological and genomic feature spaces. Furthermore, a progressive feature fusion strategy is implemented to enable effective cross-modal interaction. An RWKV-based shallow fusion module is first developed to explore cross-modal dependencies through spatial-channel hybrid operations, thereby enhancing the representational quality of fused features. A cross-RWKV deep interaction module is then designed to further strengthen information synthesis via iterative cross-attention mechanisms, while simultaneously reinforcing intra-modal representation learning and cross-modal knowledge transfer. Surv-RWKV is expected to effectively capture such cross-modal correlations, thereby improving the accuracy and interpretability of survival predictions. Extensive validation across five TCGA cancer cohorts demonstrates that Surv-RWKV achieves state-of-the-art predictive performance with superior computational efficiency.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131506"},"PeriodicalIF":7.5,"publicationDate":"2026-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192854","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":"ESG investment under different supply chain power structures: decisions, impacts, and the triple win","authors":"Man Yu ,&nbsp;Erbao Cao ,&nbsp;Yu Zhang","doi":"10.1016/j.eswa.2026.131518","DOIUrl":"10.1016/j.eswa.2026.131518","url":null,"abstract":"<div><div>ESG serves as a key global metric for corporate green and sustainable development. This study investigates a manufacturer’s investment decisions on ESG subcategories (E, S, and G) and their impacts on enterprises, consumers and the environment across different supply chain power structures. The findings show that a dominant manufacturer invests less in pollutant emission reduction and corporate governance, whereas the highest ESG investment occurs under balanced power. Under ESG investment, an enterprise’s profit increases with power, yet consumer surplus peaks under balanced power. Regarding the impacts on the environment, it depends on potential market demand. Moreover, ESG investment consistently increases the retailer’s profit and consumer surplus. When the level of social responsibility commitment is below a threshold, it can increase the manufacturer’s profit. Furthermore, this work identifies the conditions for three different types of Pareto improvement, namely, the win–win scenario for firms and consumers, the win–win scenario for consumers and the environment, and the triple win–win scenario for the three parties. This study sheds light on how enterprises make decisions regarding ESG investment to realize economic and environmental benefits.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131518"},"PeriodicalIF":7.5,"publicationDate":"2026-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191898","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":"RVFormer: Keypoint-based fusion of 4D radar and vision for 3D object detection in autonomous driving","authors":"Xin Bi ,&nbsp;Caien Weng ,&nbsp;Panpan Tong ,&nbsp;Arno Eichberger ,&nbsp;Lu Xiong","doi":"10.1016/j.eswa.2026.131497","DOIUrl":"10.1016/j.eswa.2026.131497","url":null,"abstract":"<div><div>Multi-modal fusion is crucial in autonomous driving perception, enhancing reliability, completeness, and accuracy, which extends the performance limits of perception systems. Specifically, large-scale perception through 4D radar and vision fusion has become a key research focus aimed at improving driving safety, enhancing complex scene understanding, and supporting fine-grained local planning and control. However, existing 3D object detection methods typically rely on fixed-voxel representations to maintain detection accuracy. As the perception range increases, these methods incur considerable computational overhead. While transformer-based query methods show strong potential in capturing dependencies over large receptive fields in image-domain tasks, their application in radar-vision fusion is limited due to radar point cloud sparsity and cross-modal alignment challenges. To address these limitations, we propose RVFormer, a dual-branch feature-level fusion network that uses a sparse keypoint-based query strategy to integrate features from both modalities, thereby mitigating the impact of large-scale scenes on inference speed. Additionally, we introduce clustered voxel query initialization (CVQI) to accelerate convergence and enhance object localization. By incorporating the radar voxel painter (RVP), radar-image cross-attention (RICA), and gated adaptive fusion (GAF) modules, our framework enables deep and adaptive fusion of radar and visual features, effectively mitigating issues caused by point cloud sparsity and modality inconsistency. Compared to existing radar-vision fusion models, RVFormer demonstrates competitive performance, with an inference speed of approximately 15.2 frames per second. It delivers accuracy comparable to CNN-based approaches, while outperforming baseline methods by at least 4.72% in 3D mean average precision and 5.82% in bird’s-eye view mean average precision.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"312 ","pages":"Article 131497"},"PeriodicalIF":7.5,"publicationDate":"2026-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192560","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}