Expert Systems with Applications最新文献_第9页

Assessment of renewable energy alternatives for sustainable resource policies with knowledge-based expert prioritized quantum picture fuzzy rough modelling

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-09 DOI: 10.1016/j.eswa.2025.126826

Hasan Dinçer , Serhat Yüksel , Witold Pedrycz

{"title":"Assessment of renewable energy alternatives for sustainable resource policies with knowledge-based expert prioritized quantum picture fuzzy rough modelling","authors":"Hasan Dinçer , Serhat Yüksel , Witold Pedrycz","doi":"10.1016/j.eswa.2025.126826","DOIUrl":"10.1016/j.eswa.2025.126826","url":null,"abstract":"<div><div>Renewable energy investments aim to provide the highest efficiency with limited budgets and resources. This makes it critical to make the most appropriate choice among alternatives. Hence, a priority analysis should be carried out to satisfy this missing gap in the literature. This study aims to identify the most suitable renewable energy alternative for sustainable resource policies by addressing the challenge of uncertainty in decision-making. A novel approach integrating quantum picture fuzzy rough sets, k-means clustering, and multi stepwise weight assessment ratio analysis (M−SWARA) and multi-objective optimization on the basis of ratio analysis (MOORA) methodologies is proposed. Experts are prioritized using K-means clustering based on demographic factors such as education and experience, ensuring realistic weight assignments. Quantum theory is employed to handle uncertainty and minimize information loss, while the M−SWARA method accounts for cause-and-effect relationships in weighting criteria. The findings highlight the critical importance of aligning expert prioritization and advanced fuzzy modeling for robust and adaptive decision-making in renewable energy investments. This study contributes to sustainable resource management by providing a more accurate and adaptable framework for decision-making under uncertain conditions. The integration of quantum theory and fuzzy systems minimizes uncertainty and information loss, which in turn enables more realistic results to be obtained in decision-making processes.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126826"},"PeriodicalIF":7.5,"publicationDate":"2025-02-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143420470","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

Enhanced industrial heat load forecasting in district networks via a multi-scale fusion ensemble deep learning

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126783

Zhiqiang Chen , Yu Yang , Chundi Jiang , Yi Chen , Hao Yu , Chunguang Zhou , Chuan Li

{"title":"Enhanced industrial heat load forecasting in district networks via a multi-scale fusion ensemble deep learning","authors":"Zhiqiang Chen , Yu Yang , Chundi Jiang , Yi Chen , Hao Yu , Chunguang Zhou , Chuan Li","doi":"10.1016/j.eswa.2025.126783","DOIUrl":"10.1016/j.eswa.2025.126783","url":null,"abstract":"<div><div>Accurate heating load prediction is vital for optimizing the operation of thermal systems, improving energy utilization efficiency, reducing operational costs, enhancing user satisfaction, and promoting the use of renewable energy. To facilitate short-term prediction of heat consumption in industrial areas for practical applications, a multi-scale fusion ensemble model is proposed to address the issue of pressure balance in heating networks. Specifically, (1) Hierarchical Decomposition Approach: To overcome the limitation of relying solely on historical heat load data, a hierarchical decomposition mode is designed by combining Naïve Decomposition, Empirical Mode Decomposition, and Complementary Ensemble Empirical Mode Decomposition with Adaptive Noise. This approach deeply explores the nonlinear characteristics of the heat load. (2) Integrated Heat Load Prediction Framework: An integrated prediction framework based on neural networks—including Back Propagation Networks, Recurrent Neural Networks, Long Short-Term Memory Networks, and Gated Recurrent Unit Networks is constructed. For each component, the optimal prediction model is adaptively selected, and the predicted results are fused using weighted averages. The proposed scheme was applied to 24-hour ahead heating load prediction for four regions of a thermal power company in Quzhou City, Zhejiang Province. The coefficients of determination R<sup>2</sup> achieved for the four regions were 0.8646, 0.8707, 0.8509, and 0.9422, respectively, with Mean Absolute Percentage Errors reaching 10.18%, 3.93%, 2.78%, and 2.31%. Compared with seven classical prediction models, as well as Transformer and its variants, the proposed model outperforms them across five performance indicators and demonstrates strong generalization ability.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126783"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143349230","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

Multi-level and multi-scale cross attention network of wavelet packet transform for supersonic inlet unstart prediction

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126782

Yu-Jie Wang , Yong-Ping Zhao , Yi Jin

{"title":"Multi-level and multi-scale cross attention network of wavelet packet transform for supersonic inlet unstart prediction","authors":"Yu-Jie Wang , Yong-Ping Zhao , Yi Jin","doi":"10.1016/j.eswa.2025.126782","DOIUrl":"10.1016/j.eswa.2025.126782","url":null,"abstract":"<div><div>Wavelet packet transform has demonstrated excellent performance as a time–frequency analysis method in machine fault diagnosis. However, the inherent non-linearity and multi-scale characteristics of these signals pose significant challenges to the parameter selection of wavelet packet transform. This paper introduces a novel approach by integrating attention blocks with multi-level wavelet packet transform. This integration allows for the extraction of multi-level and multi-scale features from the time–frequency domain of non-linear signals. The proposed method provides an efficient solution to enhance the feature representation of nonlinear signals through time–frequency analysis with different decomposition levels and scales. The effectiveness of this innovative approach is validated through experiments conducted on a real inlet model dataset. The results demonstrate that the proposed model achieved comprehensive testing accuracy of 99.48% and 99.36% in two experimental schemes, respectively, indicating that its recognition performance is superior to other existing supersonic inlet unstart prediction methods.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126782"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143348833","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

Target speaker lipreading by audio–visual self-distillation pretraining and speaker adaptation

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126741

Jing-Xuan Zhang , Tingzhi Mao , Longjiang Guo , Jin Li , Lichen Zhang

{"title":"Target speaker lipreading by audio–visual self-distillation pretraining and speaker adaptation","authors":"Jing-Xuan Zhang , Tingzhi Mao , Longjiang Guo , Jin Li , Lichen Zhang","doi":"10.1016/j.eswa.2025.126741","DOIUrl":"10.1016/j.eswa.2025.126741","url":null,"abstract":"<div><div>Lipreading is an important technique for facilitating human–computer interaction in noisy environments. Our previously developed self-supervised learning method, AV2vec, which leverages multimodal self-distillation, has demonstrated promising performance in speaker-independent lipreading on the English LRS3 dataset. However, AV2vec faces challenges such as high training costs and a potential scarcity of audio–visual data for lipreading in languages other than English, such as Chinese. Additionally, most studies concentrate on speaker-independent lipreading models, which struggle to account for the substantial variation in speaking styles across different speakers. To address these issues, we propose a comprehensive approach. First, we investigate cross-lingual transfer learning, adapting a pre-trained AV2vec model from a source language and optimizing it for the lipreading task in a target language. Second, we enhance the accuracy of lipreading for specific target speakers through a speaker adaptation strategy, which is not extensively explored in previous research. Third, after analyzing the complementary performance of lipreading with lip region-of-interest (ROI) and face inputs, we introduce a model ensembling strategy that integrates both, significantly boosting model performance. Our method achieved a character error rate (CER) of 77.3% on the evaluation set of the ChatCLR dataset, which is lower than the top result from the 2024 Chat-scenario Chinese Lipreading Challenge.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126741"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143379406","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

Learning unified distance metric for heterogeneous attribute data clustering

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126738

Yiqun Zhang , Mingjie Zhao , Yizhou Chen , Yang Lu , Yiu-ming Cheung

{"title":"Learning unified distance metric for heterogeneous attribute data clustering","authors":"Yiqun Zhang , Mingjie Zhao , Yizhou Chen , Yang Lu , Yiu-ming Cheung","doi":"10.1016/j.eswa.2025.126738","DOIUrl":"10.1016/j.eswa.2025.126738","url":null,"abstract":"<div><div>Datasets composed of numerical and categorical attributes (also called mixed data hereinafter) are common in real clustering tasks. Differing from numerical attributes that indicate tendencies between two concepts (e.g., high and low temperature) with their values in well-defined Euclidean distance space, categorical attribute values are different concepts (e.g., different occupations) embedded in an implicit space. Simultaneously exploiting these two very different types of information is an unavoidable but challenging problem, and most advanced attempts either encode the heterogeneous numerical and categorical attributes into one type, or define a unified metric for them for mixed data clustering, leaving their inherent connection unrevealed. This paper, therefore, studies the connection among any-type of attributes and proposes a novel Heterogeneous Attribute Reconstruction and Representation (HARR) learning paradigm accordingly for cluster analysis. The paradigm transforms heterogeneous attributes into a homogeneous status for distance metric learning, and integrates the learning with clustering to automatically adapt the metric to different clustering tasks. Differing from most existing works that directly adopt defined distance metrics or learn attribute weights to search clusters in a subspace. We propose to project the values of each attribute into unified learnable multiple spaces to more finely represent and learn the distance metric for categorical data. HARR is parameter-free, convergence-guaranteed, and can more effectively self-adapt to different sought number of clusters <span><math><mi>k</mi></math></span>. Extensive experiments illustrate its superiority in terms of accuracy and efficiency.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126738"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143403070","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

LRM-MVSR: A lightweight birdsong recognition model based on multi-view feature extraction enhancement and spatial relationship capture

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126735

Jing Wan , Zhongxiang Lin , Zhiqi Zhu , Wanhong Yang , Aibin Chen , Yurong Sun

{"title":"LRM-MVSR: A lightweight birdsong recognition model based on multi-view feature extraction enhancement and spatial relationship capture","authors":"Jing Wan , Zhongxiang Lin , Zhiqi Zhu , Wanhong Yang , Aibin Chen , Yurong Sun","doi":"10.1016/j.eswa.2025.126735","DOIUrl":"10.1016/j.eswa.2025.126735","url":null,"abstract":"<div><div>Bird diversity is crucial to biodiversity, and bird song recognition plays a key role in population monitoring and ecological conservation. However, bird song recognition faces two major challenges in practical applications: first, background noise interference affects the spatial representation of features, and second, the model is computationally expensive. To address these problems, this paper proposes a lightweight birdsong recognition method (LRM-MVSR) based on multi-view feature extraction enhancement and spatial relationship capture. The method first extracts bird song features using Log Power Spectrum (LPS) and Meier frequency cepstrum coefficients (MFCC), then enhances the expressiveness of bird song signal features by suppressing the background noise using Multi-View Feature Extraction Enhancement (MVFE), and further distinguishes the spatial relationships among different features using Pyramid Split Residual Block (PSRB). To reduce the complexity of the model, ShuffleNetV1 downsampling module and Shuffle Attention are also integrated to construct the backbone network SANet, which enhances the model’s ability to perceive different bird song sound features and effectively discriminate the similarity of features in bird songs. The experimental results show that LRM-MVSR achieves recognition accuracies of 97.84%, 95.71%, and 94.80% on the customized Birdselfdata, public Urbansound 8K, and Birdsdata datasets, respectively, which are superior to existing methods. These results indicate that LRM-MVSR effectively solves the key challenges in bird song recognition, while realizing efficient and accurate recognition, providing a powerful tool for ecological conservation.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126735"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143377204","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

Linearized circular energy curve color image segmentation based on Tsallis entropy

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126746

Shaoxun Wang , Jiulun Fan , Heng Liu

{"title":"Linearized circular energy curve color image segmentation based on Tsallis entropy","authors":"Shaoxun Wang , Jiulun Fan , Heng Liu","doi":"10.1016/j.eswa.2025.126746","DOIUrl":"10.1016/j.eswa.2025.126746","url":null,"abstract":"<div><div>Image processing is a continuously evolving field, and due to the fact that color images contain more information than grayscale images and are the most encountered type of images in daily life, the analysis and processing of color images are particularly important. The use of the H component of the HSI color space model to construct a circular histogram for selecting the optimal threshold is one method in color image segmentation. However, traditional circular histograms only reflect the pixel distribution of the entire image by statistically counting the frequency of pixel occurrences in the entire image, without considering the spatial context information of the image for optimal threshold selection. To address this drawback, we propose the method of circular energy curves (CEC). This method involves calculating energy curves on the H component to construct CEC. The optimal threshold is then selected based on these curves. To reduce the complexity of threshold selection for CEC and enhance their generality, we combine cumulative distribution functions with information energy, introducing the method of cumulative distribution information energy (CDF-IE) to determine the optimal breakpoint. After linearizing the circular energy curves (L-CEC), we use the Tsallis entropy thresholding method to achieve the optimal segmentation threshold selection for the H component CEC. To improve the effectiveness of the algorithm, we present the Tsallis entropy recursive algorithm after L-CEC, which reduces the time complexity. The proposed method in this paper, compared to the circular histogram method, achieves better image segmentation results and exhibits stronger applicability.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"273 ","pages":"Article 126746"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143454410","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

Cross-city traffic state prediction based on knowledge transfer framework

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126747

Dongwei Xu, Yufu Tang, Jingfei Ju, Zefeng Yu, Jiaying Zheng, Tongcheng Gu, Haifeng Guo

{"title":"Cross-city traffic state prediction based on knowledge transfer framework","authors":"Dongwei Xu, Yufu Tang, Jingfei Ju, Zefeng Yu, Jiaying Zheng, Tongcheng Gu, Haifeng Guo","doi":"10.1016/j.eswa.2025.126747","DOIUrl":"10.1016/j.eswa.2025.126747","url":null,"abstract":"<div><div>Real-time and accurate traffic state prediction is crucial for smart transportation systems in cities, serving as the foundation for various downstream intelligent mobility applications. However, existing traffic prediction models heavily rely on large-scale traffic data, which require significant cost expenses. Therefore, we propose a novel knowledge transfer framework to achieve accurate prediction of cross-city traffic state and to reduce expenses. Firstly, a feature extraction module based on a shared processing mechanism is proposed to achieve the intrinsic connection of multi-graph networks, which uses multi-layers graph convolution and instance attention layer to extract embedding feature. Secondly, a feature matching module based on Linear Transformer is proposed to achieve feature adaptation between the source city and the target city, which utilizes self-attention and interaction attention mechanism to match and generalize the embedding feature. Then, a joint meta-learning module is used for the pre-training of traffic prediction model, which includes source training, target fine-tuning, and parameters updating. Finally, the final fine-tuning of the traffic prediction model is carried out on target city to achieve accurate prediction of cross-city traffic state. Experimental results conducted on real traffic datasets demonstrate that the model framework proposed in this paper outperforms baseline models in terms of performance.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126747"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143377074","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

Modeling iot traffic patterns: Insights from a statistical analysis of an mtc dataset

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126726

David E. Ruiz-Guirola , Onel L.A. López , Samuel Montejo-Sánchez

{"title":"Modeling iot traffic patterns: Insights from a statistical analysis of an mtc dataset","authors":"David E. Ruiz-Guirola , Onel L.A. López , Samuel Montejo-Sánchez","doi":"10.1016/j.eswa.2025.126726","DOIUrl":"10.1016/j.eswa.2025.126726","url":null,"abstract":"<div><div>The Internet-of-Things (IoT) is rapidly expanding, connecting numerous devices and becoming integral to our daily lives. As this occurs, ensuring efficient traffic management becomes crucial. Effective IoT traffic management requires modeling and predicting intricate machine-type communication (MTC) dynamics, for which machine-learning (ML) techniques are certainly appealing. However, obtaining comprehensive and high-quality datasets, along with accessible platforms for reproducing ML-based predictions, continues to impede the research progress. In this paper, we aim to fill this gap by characterizing the Smart Campus MTC dataset provided by the University of Oulu. Specifically, we perform a comprehensive statistical analysis of the MTC traffic utilizing goodness-of-fit tests, including well-established tests such as Kolmogorov–Smirnov, Anderson–Darling, chi-squared and root mean square error. The analysis centers on examining and evaluating three models that accurately represent the two most significant MTC traffic types: periodic updating and event-driven, which are also identified from the dataset. The results demonstrate that the models accurately characterize the traffic patterns. The Poisson point process model exhibits the best fit for event-driven patterns with errors below 11%, while the quasi-periodic model fits accurately the periodic updating traffic with errors below 7%.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126726"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143349229","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"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}

引用次数: 0

Label-guided diversified learning model for occluded person re-identification

IF 7.5 1区计算机科学

Expert Systems with Applications Pub Date : 2025-02-08 DOI: 10.1016/j.eswa.2025.126745

Lun Zhang, Shuli Cheng , Liejun Wang

{"title":"Label-guided diversified learning model for occluded person re-identification","authors":"Lun Zhang, Shuli Cheng , Liejun Wang","doi":"10.1016/j.eswa.2025.126745","DOIUrl":"10.1016/j.eswa.2025.126745","url":null,"abstract":"<div><div>In the field of person re-identification (Re-ID), the challenge of occlusion remains formidable. Whether it is object or person occlusion, these factors disrupt the extraction of target features by models, leading to insufficient perception and learning of target features across diverse occlusion scenarios. Current methods often address this by using fixed positioning or auxiliary networks to segment features for evaluating visible target features. However, these approaches suffer from low accuracy or efficiency, failing to fully capture comprehensive target features. To tackle these issues, this paper proposes a novel Label-Guided Diversified Learning (LGDL) model. Specifically, the Occlusion Simulation based on Label Guidance (OSBLG) module employs two occlusion simulation processes tailored for person and object occlusion, mitigating their impact on model performance. This module implicitly guides the model to perceive target and occluded regions using identity labels in simulated images. Subsequently, the Feature Diversification Learning (FDL) module enhances adaptive learning of target features by extending class token as well as constraining multi-head features, and uses a Dynamic Weighted Dual Constraint (DWDC) loss to enforce the model’s learning of distinct features, thereby achieving diversified feature learning in target regions. Extensive experiments demonstrate that the proposed LGDL model outperforms state-of-the-art methods on occluded, partial, and holistic Re-ID datasets. Particularly, on the Occluded-DukeMTMC dataset, LGDL achieves performance metrics of 67.5% mAP score and 78.2% Rank-1 accuracy.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"272 ","pages":"Article 126745"},"PeriodicalIF":7.5,"publicationDate":"2025-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143379262","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