Expert Systems with Applications最新文献_第4页

Beyond-Skeleton: Zero-shot Skeleton Action Recognition enhanced by supplementary RGB visual information 超越骨架：通过补充 RGB 视觉信息增强零镜头骨架动作识别能力

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-15 DOI: 10.1016/j.eswa.2025.126814

Hongjie Liu , Yingchun Niu , Kun Zeng , Chun Liu , Mengjie Hu , Qing Song

{"title":"Beyond-Skeleton: Zero-shot Skeleton Action Recognition enhanced by supplementary RGB visual information","authors":"Hongjie Liu , Yingchun Niu , Kun Zeng , Chun Liu , Mengjie Hu , Qing Song","doi":"10.1016/j.eswa.2025.126814","DOIUrl":"10.1016/j.eswa.2025.126814","url":null,"abstract":"<div><div>Zero-shot action recognition (ZSAR) recognizes action categories that have not appeared during the training process and has garnered widespread attention due to its potential to save costs in retraining and data annotation. We observed that the existing ZSAR method based on skeleton sequences only uses human posture information in the skeleton sequence, lacks discriminative semantic representation in some similar behavior recognition, and lacks effective interaction between different modalities, resulting in unsatisfactory performance and limited applications of the ZSAR. To solve these problems, we propose a novel method, called Beyond-Skeleton zero-shot Learning (BSZSL), which is used to enhance zero-shot Skeleton Action Recognition. Firstly, a multi-prompt learning strategy is introduced. It utilizes prompt information to guide the model to simultaneously learn complementary semantic information related to behavior categories from both skeleton sequences and RGB information, making the visual feature information more expressive. Specifically, it employs a pre-trained multimodal model to extract prior knowledge related to behaviors from RGB and then guides the skeleton sequence features using this knowledge. This enhances the complementary features of both RGB and skeleton modalities. Secondly, to constrain the mapping relationship of different modal feature information, a Contrastive Clustering (CC) module is designed. This module emphasizes the similarity of features within the same category while increasing the differences in feature mapping between different categories. Finally, evaluating our method on the NTU-60 and NTU-120 datasets with multi-split settings, the result demonstrates that our method achieves state-of-the-art performance in both zero-shot learning (ZSL) and generalized zero-shot learning (GZSL) settings.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126814"},"PeriodicalIF":7.5,"publicationDate":"2025-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143446057","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}

引用次数: 0

ODTE—An ensemble of multi-class SVM-based oblique decision trees

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-15 DOI: 10.1016/j.eswa.2025.126833

Ricardo Montañana, José A. Gámez, José M. Puerta

{"title":"ODTE—An ensemble of multi-class SVM-based oblique decision trees","authors":"Ricardo Montañana, José A. Gámez, José M. Puerta","doi":"10.1016/j.eswa.2025.126833","DOIUrl":"10.1016/j.eswa.2025.126833","url":null,"abstract":"<div><div>We propose ODTE, a new ensemble that uses oblique decision trees as base classifiers. Additionally, we introduce STree, the base algorithm for growing oblique decision trees, which leverages support vector machines to define hyperplanes within the decision nodes. We embed a multiclass strategy (one-vs-one or one-vs-rest) at the decision nodes, allowing the model to directly handle non-binary classification tasks without the need to cluster instances into two groups, as is common in other approaches from the literature. In each decision node, only the best-performing model (SVM)—the one that minimizes an impurity measure for the n-ary classification—is retained, even if the learned SVM addresses a binary classification subtask. An extensive experimental study involving 49 datasets and various state-of-the-art algorithms for oblique decision tree ensembles has been conducted. Our results show that ODTE ranks consistently above its competitors, achieving significant performance gains when hyperparameters are carefully tuned. Moreover, the oblique decision trees learned through STree are more compact than those produced by other algorithms evaluated in our experiments.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126833"},"PeriodicalIF":7.5,"publicationDate":"2025-02-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143429279","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}

引用次数: 0

Joint optimization of quality control and maintenance policy for a production system with quality-dependent failures

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-14 DOI: 10.1016/j.eswa.2025.126800

Jingjing Wang , Lingyun Luo , Guoqing Mu , Yingying Ma , Chao Ni

{"title":"Joint optimization of quality control and maintenance policy for a production system with quality-dependent failures","authors":"Jingjing Wang , Lingyun Luo , Guoqing Mu , Yingying Ma , Chao Ni","doi":"10.1016/j.eswa.2025.126800","DOIUrl":"10.1016/j.eswa.2025.126800","url":null,"abstract":"<div><div>A machine is either subject to hard failure or soft failure, while quality-dependent failure is usually ignored in production systems. However, in practice, non-conforming products generally accelerate the degradation process of production systems. To fill these gaps, this paper formulated an integrated model of the optimal quality control policy and maintenance policies under the quality-dependent failures for production systems. Since the severe degradation of product machines directly impacts the non-conforming rate, effective preventive and opportunistic maintenance actions are necessary. Moreover, the production system can timely be corrected from an out-of-control state to a control state after adopting a minimal repair. An opportunity is created based on the buffer inventory level instead of the machine degradation level, which is different from previous research. The production machine can be opportunistically maintained at the point of maximum inventory. The renewal process is utilized to derive the integrated optimization model, and the collaboration relationship among production, quality control and maintenance management problems is taken into consideration. The system’s total cost rate is minimized by optimizing the preventive maintenance threshold and sampling control coefficients. Numerical examples are given to illustrate the priority of the proposed model and the optimal results are obtained by differential evaluation algorithm, which provides a more meaningful perspective for managers.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126800"},"PeriodicalIF":7.5,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143403699","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}

引用次数: 0

Graph Attention Networks For Anomalous Drone Detection: RSSI-Based Approach with Real-world Validation

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-14 DOI: 10.1016/j.eswa.2025.126913

Ghulam E Mustafa Abro , Ayman M Abdallah

{"title":"Graph Attention Networks For Anomalous Drone Detection: RSSI-Based Approach with Real-world Validation","authors":"Ghulam E Mustafa Abro , Ayman M Abdallah","doi":"10.1016/j.eswa.2025.126913","DOIUrl":"10.1016/j.eswa.2025.126913","url":null,"abstract":"<div><div>The swift proliferation of unmanned aerial vehicles (UAVs) and their expanding applications have engendered considerable security apprehensions, especially with the detection of anomalous drones inside swarms. This research introduces an innovative methodology utilising Graph Attention Networks (GAT) and Received Signal Strength Indicator (RSSI) data to discover and identify abnormal drones in UAV networks. The suggested method employs a V-cycle algorithm-based graph attention model, wherein RSSI deviations from the mean are calculated for each drone node and utilised as a feature within the graph. A radius graph is created to illustrate drone-to-drone conversations, facilitating the computation of attention scores that assess the significance of each node’s connectivity and RSSI attributes. Drones displaying irregular RSSI patterns, as detected by the GAT framework, are identified as potential dangers or anomalous drones. The system is engineered to manage intricate real-world settings by effectively detecting drones exhibiting aberrant behaviour via multilevel graph coarsening and refinement methodologies. To assess the efficacy of the suggested strategy, simulations were executed, and empirical experiments were carried out with the Robolink Codrones kit. The trials validated the system’s capability to identify drones exhibiting anomalous signal strength fluctuations in real-time situations. The findings illustrate the suggested method’s efficacy in detecting anomalous drones using RSSI anomalies, surpassing conventional detection techniques in accuracy and computing efficiency. RSSI data and graph attention approaches for autonomous drone identification can improve UAV network security and anomaly detection systems, as shown in this study.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126913"},"PeriodicalIF":7.5,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143438163","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}

引用次数: 0

CILOSR: A unified framework for enhanced class incremental learning based open-set human activity recognition using wearable sensors

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-14 DOI: 10.1016/j.eswa.2025.126893

Cheng Wang , Lin Chen , Bangwen Zhou , Yaqiao Xian , Yuhao Zhao , Zhan Huan

{"title":"CILOSR: A unified framework for enhanced class incremental learning based open-set human activity recognition using wearable sensors","authors":"Cheng Wang , Lin Chen , Bangwen Zhou , Yaqiao Xian , Yuhao Zhao , Zhan Huan","doi":"10.1016/j.eswa.2025.126893","DOIUrl":"10.1016/j.eswa.2025.126893","url":null,"abstract":"<div><div>The field of Human Activity Recognition (HAR) has seen widespread adoption of wearable sensors for the collection of time-series signals. However, as new activities emerge, HAR systems struggle to differentiate novel categories from existing ones, as they are trained on a fixed set of known classes. To overcome this limitation, an innovative framework called CILOSR is designed for the continuous integration of novel, previously unseen activity classes into HAR models. The proposed CILOSR framework combines two pivotal processes, Class Incremental Learning (CIL) to enhance model knowledge with newly acquired data, while Open-Set Recognition (OSR) to detect and characterize new activity classes. The CIL phase employs extreme point updating based Extreme Value Machine algorithm, which preserves and updates the reference boundary points and extreme value vectors for established classes alongside new data integration. For the OSR phase, Principal Component Analysis (PCA) is incorporate to reduce feature redundancy within the time–frequency domain, thereby refining the feature space. Subsequently, Particle Swarm Optimization (PSO) is utilized for precise calibration of Extreme Value Machine (EVM) parameters to optimize the recognition process. Several experiments on the UCI, PAMAP2, and USC-HAD datasets confirm the effectiveness of the CILOSR framework. Specifically, OSR-LPC (Leave-Partial-Class) experiments on the UCI dataset demonstrate that CILOSR with PSO-EVM (Cosine) + PCA significantly outperforms the standard EVM (Cosine). The model achieves F1-macro score of 0.88 and accuracy of 0.89, compared to the baseline’s 0.59 and 0.66. These results highlight CILOSR’s enhanced accuracy in recognizing both known and unknown activities, demonstrating its potential for dynamic and scalable HAR applications.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126893"},"PeriodicalIF":7.5,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143453670","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}

引用次数: 0

Inspired by “Focus, Fusion, Collaboration”: A multi-level ensemble network for automatic pneumonia diagnosis from full slice CT images

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-14 DOI: 10.1016/j.eswa.2025.126806

Linna Zhao, Jianqiang Li, Qing Zhao, Xi Xu

{"title":"Inspired by “Focus, Fusion, Collaboration”: A multi-level ensemble network for automatic pneumonia diagnosis from full slice CT images","authors":"Linna Zhao, Jianqiang Li, Qing Zhao, Xi Xu","doi":"10.1016/j.eswa.2025.126806","DOIUrl":"10.1016/j.eswa.2025.126806","url":null,"abstract":"<div><div>Pneumonia is an infectious disease that endangers human health. With advancements in science and technology, deep learning-driven techniques have gained prominence in this field. However, their applicability to clinical practice remains limited because they mostly neglect three key points: focus on local lesion regions, multi-level feature fusion, and sequential collaborative decision-making. In this paper, we present a novel multi-level ensemble network for automatic pneumonia diagnosis from full slice CT images, inspired by the “Focus, Fusion, Collaboration” strategy. Our proposed model involves three modules: the global–local feature extraction module is first designed to fully extract the global structure information and local lesion details; subsequently, the multi-level feature fusion module is responsible for integrating the above-mentioned global and local information; finally, the sequential pneumonia prediction module is utilized to learn the contextual relationship between the adjacent slices, thus generating the final diagnosis results. Building upon mimicking the diagnostic behavior from real-world clinical scenarios, our model enables the integration of multiple types of information (including global structure information, local lesion features, and slice dependencies) and sequential pneumonia diagnosis. Extensive comparative experiments are conducted to verify the feasibility and effectiveness of our proposed method. The experimental results show that our model can obtain an accuracy of 91.4% in a four-class pneumonia diagnosis task, outperforming the other classical works.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126806"},"PeriodicalIF":7.5,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143428996","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}

引用次数: 0

ShadowAdapter: Adapting Segment Anything Model with Auto-Prompt for shadow detection

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-14 DOI: 10.1016/j.eswa.2025.126809

Leiping Jie , Hui Zhang

引用次数: 0

Multi-teacher knowledge distillation for debiasing recommendation with uniform data

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-14 DOI: 10.1016/j.eswa.2025.126808

Xinxin Yang, Xinwei Li, Zhen Liu, Yafan Yuan, Yannan Wang

{"title":"Multi-teacher knowledge distillation for debiasing recommendation with uniform data","authors":"Xinxin Yang, Xinwei Li, Zhen Liu, Yafan Yuan, Yannan Wang","doi":"10.1016/j.eswa.2025.126808","DOIUrl":"10.1016/j.eswa.2025.126808","url":null,"abstract":"<div><div>Recent studies have highlighted the bias problem in recommender systems which affects the learning of users’ true preferences. One significant reason for bias is that the training data is missing not at random (MNAR). While existing approaches have demonstrated the usefulness of uniform data that is missing at random (MAR) for debiasing, the current models lack a comprehensive exploration of unbiased features within uniform data. Considering the valuableness and limited size of uniform data, this paper proposes a multi-teacher knowledge distillation framework (UKDRec) to extract and transfer more unbiased information from uniform data. The proposed framework consists of two components: a label-based teacher model that leverages supervision signals and a feature-based teacher model that facilitates the transfer of comprehensive unbiased features. To effectively extract unbiased features, we introduce a contrastive learning strategy that combines the uniform data with control data. The framework is trained using a multi-task learning approach, which enhances the transfer of unbiased knowledge. Extensive experiments conducted on real-world datasets demonstrate the superior debiasing performance of our approach compared to competitive baselines.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126808"},"PeriodicalIF":7.5,"publicationDate":"2025-02-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143438083","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}

引用次数: 0

Human centric VR system development supporting fire emergency evacuation: A novel knowledge-data dual driven approach 支持火灾紧急疏散的以人为中心的虚拟现实系统开发：知识-数据双驱动的新方法

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-13 DOI: 10.1016/j.eswa.2025.126895

Jiaxin Ling , Xiaojun Li , Yi Shen , Chao Chen , Zhiguo Yan , Hehua Zhu , Haijiang Li

{"title":"Human centric VR system development supporting fire emergency evacuation: A novel knowledge-data dual driven approach","authors":"Jiaxin Ling , Xiaojun Li , Yi Shen , Chao Chen , Zhiguo Yan , Hehua Zhu , Haijiang Li","doi":"10.1016/j.eswa.2025.126895","DOIUrl":"10.1016/j.eswa.2025.126895","url":null,"abstract":"<div><div>Catastrophic fire accidents happened inside the tunnel have made it evident that human factors, especially misconduct, should be taken into account when it comes to fire emergency evacuation. However, conventional approaches separate fire safety education from evacuation training, failing to account for individual capabilities and behavioral dynamics, resulting in less intuitive and ineffective preparedness. A human-centric and more adaptive training for tunnel fire evacuation which takes both knowledge learning and behavior training into account is in urgent need. Motivated by such need, this study proposes a knowledge-data dual driven (KD3) framework, to seamlessly combine tunnel fire knowledge transfer and evacuation training into a unified system. A Virtual Reality (VR) system is developed based on KD3, which is composed of interactive fire-knowledge transfer module and immersive fire training module. To verify the applicability and effectiveness of the established system, the interactive fire-knowledge transfer module was open to public for different tunnel users to learn, and a total of 50 participants were recruited to conduct VR training. Results verify the rationale of the developed system, as well as the proposed KD3 framework, demonstrating that the integration of knowledge learning and VR training significantly improves individuals’ evacuation decision-making and escape behavior during tunnel fires. These findings contribute to a paradigm shift in fire evacuation training by bridging the gap between theoretical learning and practical application. The study provides critical insights into human-centric emergency preparedness and offers practical guidance for future adaptive training systems in emergency.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126895"},"PeriodicalIF":7.5,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143429278","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}

引用次数: 0

A high-precision and lightweight ore particle segmentation network for industrial conveyor belt

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-13 DOI: 10.1016/j.eswa.2025.126891

Hanquan Zhang , Dong Xiao

{"title":"A high-precision and lightweight ore particle segmentation network for industrial conveyor belt","authors":"Hanquan Zhang , Dong Xiao","doi":"10.1016/j.eswa.2025.126891","DOIUrl":"10.1016/j.eswa.2025.126891","url":null,"abstract":"<div><div>The existing industrial ore particle segmentation methods are challenged by particle size, different shapes and textures, dust shielding and lack of field data, so segmentation accuracy and efficiency are low. To alleviate the above problems faced by ore image segmentation (OIS), a novel end-to-end ore image segmentation network (OIS-Net) is proposed in this paper. First, we propose for the first time a lightweight residual architecture called RDSB to improve speed of ore particle segmentation by OIS-Net. Secondly, we also propose a novel channel and spatial sequence fusion mechanism (CSSFM) to enhance the network’s ability to accurately extract and understand complex ore features, which is beneficial to improve the segmentation accuracy of ore images. And then, improve and apply atrous spatial pyramid pooling (ASPP+) module to improve multi-scale feature fusion capability for network, which is conducive to identification and segmentation of small-scale ores. Finally, we went to industrial site to collect and construct a large, high-quality conveyor belt ore dataset Industrial-Ore 700 using high-performance image sensors, which solved the problem regarding unavailability for public datasets. After extensive experiments, it is proved that OIS-Net is superior to common segmentation methods in both performance and speed, among which segmentation accuracy of OIS-Net is as high as 96% and segmentation time is the shortest.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126891"},"PeriodicalIF":7.5,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143453659","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}

引用次数: 0