Applied Soft Computing最新文献

{"title":"Digital twin-assisted fault diagnosis framework for rolling bearings under imbalanced data","authors":"Zhen Ming,&nbsp;Baoping Tang,&nbsp;Lei Deng,&nbsp;Qichao Yang,&nbsp;Qikang Li","doi":"10.1016/j.asoc.2024.112528","DOIUrl":"10.1016/j.asoc.2024.112528","url":null,"abstract":"<div><div>The application of deep learning-based fault diagnosis methods is constrained by the imbalanced data. Recently, many studies have suggested integrating dynamic model responses into the training process to address data imbalances. However, significant distribution discrepancies exist between dynamic model responses and real measured data, resulting in suboptimal performance. To address this challenge, this research proposes a digital twin-assisted framework for rolling bearings fault diagnosis under imbalanced data, which minimizes the distribution discrepancies between dynamic model responses and real measured data through information and feature transfer. Firstly, a Digital Twin-assisted Data Fusion Strategy (DTDFS) is proposed to facilitate information transfer from physical entities to dynamic models, generating digital twin data for data augmentation. Subsequently, a Frequency Filter Subdomain Adaptation Network (FFSAN) is proposed to achieve feature transfer between twin data and measured data. Finally, experimental results and engineering applications demonstrate that the proposed framework significantly outperforms existing imbalanced fault diagnosis methods, which is crucial to the application of deep learning-based fault diagnosis in industrial settings.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112528"},"PeriodicalIF":7.2,"publicationDate":"2024-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745498","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":"Enhancing the performance of CNN models for pneumonia and skin cancer detection using novel fractional activation function","authors":"Meshach Kumar,&nbsp;Utkal Mehta","doi":"10.1016/j.asoc.2024.112500","DOIUrl":"10.1016/j.asoc.2024.112500","url":null,"abstract":"<div><div>This paper introduces a novel Riemann–Liouville (RL) conformable fractional derivative based Adaptable-Shifted-Fractional-Rectified-Linear-Unit, briefly called <span><math><mrow><msup><mrow></mrow><mrow><mi>R</mi><mi>L</mi></mrow></msup><mi>ASFReLU</mi></mrow></math></span>, and evaluates its efficacy in enhancing the performance of convolutional neural network (CNN) models for pneumonia and skin cancer detection. The study conducts a comprehensive comparative analysis against traditional activation functions and state-of-the-art CNN architectures. The results show that <span><math><mrow><msup><mrow></mrow><mrow><mi>R</mi><mi>L</mi></mrow></msup><mi>ASFReLU</mi></mrow></math></span> consistently outperforms other functions, achieving higher accuracy. Comparative evaluations with various neural network architectures reveal that the model equipped with <span><math><mrow><msup><mrow></mrow><mrow><mi>R</mi><mi>L</mi></mrow></msup><mi>ASFReLU</mi></mrow></math></span> exhibits superior performance despite its simplicity and fewer trainable parameters, highlighting its efficiency and effectiveness. The findings suggest that <span><math><mrow><msup><mrow></mrow><mrow><mi>R</mi><mi>L</mi></mrow></msup><mi>ASFReLU</mi></mrow></math></span> holds promise in improving diagnostic accuracy and efficiency in medical imaging applications, contributing to advancements in healthcare technology and facilitating better patient care. The proposed fractional nonlinear transformation can offer high performance with reduced computational cost, making it practical for deployment in healthcare settings.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112500"},"PeriodicalIF":7.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745489","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":"Hierarchical Scale Awareness for object detection in Unmanned Aerial Vehicle Scenes","authors":"Shijie Wang ,&nbsp;Chaoying Wan ,&nbsp;Jinqiang Yan ,&nbsp;Silong Li ,&nbsp;Tianmeng Sun ,&nbsp;Jieru Chi ,&nbsp;Guowei Yang ,&nbsp;Chenglizhao Chen ,&nbsp;Teng Yu","doi":"10.1016/j.asoc.2024.112487","DOIUrl":"10.1016/j.asoc.2024.112487","url":null,"abstract":"<div><div>Existing object detection models are typically designed without considering the small-scale context, leading to significant challenges in detecting small objects within Unmanned Aerial Vehicle (UAV) scenes. Therefore, this paper aims to incorporate a novel hierarchical scale-aware module into the neck component of the classical YOLO architecture. This module hierarchically enhances the object features, progressing from small to large scales. Specifically, the proposed Small-Scale Awareness (SSA) module is designed to enhance features from small-scale objects, while the introduced Receptive Field Expansion (RFE) module is responsible for modeling contextual information in a way that expands the receptive field while maintaining feature diversity for large-scale objects. Additionally, in the backbone of our model, a Stack of Non-Linear Mapping (SNM) module is proposed, which utilizes deformable convolutions to fuse feature maps of diverse scales through a cascade of non-linear mapping units, to capture a wide range of contextual and discriminative information. The experimental results on the VisDrone dataset demonstrate that the proposed model outperforms the state-of-the-art models both on the mean Average Precision (mAP) and Average Precision 50 (AP50) metrics. The ablation studies have proved that the proposed modules are beneficial to improve the detection performance of objects in UAV scenes.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112487"},"PeriodicalIF":7.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745488","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":"Correcting translation for non-autoregressive transformer","authors":"Shuheng Wang ,&nbsp;Heyan Huang ,&nbsp;Shumin Shi ,&nbsp;Dongbai Li ,&nbsp;Dongen Guo","doi":"10.1016/j.asoc.2024.112488","DOIUrl":"10.1016/j.asoc.2024.112488","url":null,"abstract":"<div><div>Non-Autoregressive Transformer has shown great success in recent years. It generally employs the encoder–decoder framework, where the encoder maps the sentence into hidden representation, and the decoder generates the target tokens simultaneously. Since the theory of non-autoregressive transformer is consistent with the architecture of the encoder, we suppose that it is somewhat wasteful for the encoder to only map input sentence into hidden representation. In this study, we proposed a novel non-autoregressive transformer to fully exploit the capabilities of the encoder. Specifically, in our approach, the encoder not only encodes the input sentence into hidden representation, but also generates the target tokens. Consequently, the decoder is relieved of its responsibility to generate the target tokens, instead of focusing on correcting the sentence produced by the encoder. We evaluate the performance of the proposed non-autoregressive transformer on three widely-used translation tasks. The experimental results illustrate the proposed method can significantly improve the performance of the non-autoregressive transformer , which achieved 27.94 BLEU on WMT14 EN <span><math><mo>→</mo></math></span> DE task, 33.96 BLEU on WMT16 EN <span><math><mo>→</mo></math></span> RO task, and 33.85 BLEU on IWSLT14 DE <span><math><mo>→</mo></math></span> EN.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112488"},"PeriodicalIF":7.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745494","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":"Anomaly detection and segmentation in industrial images using multi-scale reverse distillation","authors":"Chien-Liang Liu,&nbsp;Chia-Chen Chung","doi":"10.1016/j.asoc.2024.112502","DOIUrl":"10.1016/j.asoc.2024.112502","url":null,"abstract":"<div><div>Anomaly detection and segmentation in industrial images are critical tasks requiring robust and precise methodologies. This paper presents the Multi-Scale Reverse Distillation (MSRD) methodology, an innovative improvement of the foundational reverse distillation approach. MSRD leverages autoencoder-based techniques integrated with information at different levels to significantly enhance reconstruction capabilities. A novel module incorporated at the decoder’s end facilitates precise sample reconstruction. The proposed loss function incorporates the reconstruction loss <span><math><msub><mrow><mi>L</mi></mrow><mrow><mi>R</mi><mi>e</mi><mi>c</mi><mi>o</mi><mi>n</mi></mrow></msub></math></span>, calculated using structural similarity index measure (SSIM) between the original and reconstructed images, in addition to the knowledge distillation loss <span><math><msub><mrow><mi>L</mi></mrow><mrow><mi>K</mi><mi>D</mi></mrow></msub></math></span>. Additionally, the integration of a feature pyramid network improves the spatial coherence of anomaly maps across varying scales, enabling detailed anomaly segmentation. The MSRD method undergoes rigorous evaluation on three public datasets, demonstrating superior performance in both anomaly detection and segmentation. The results highlight MSRD’s adaptability and effectiveness in one-class learning-based applications. This study underscores MSRD’s potential as a powerful tool for industrial anomaly detection, offering significant advancements in AI-driven image analysis.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112502"},"PeriodicalIF":7.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745491","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":"Trainable Monte Carlo-MLP for cost uncertainty in resilient supply chain optimization with additive manufacturing implementation challenges","authors":"Pardis Roozkhosh,&nbsp;Mojtaba Ghorbani","doi":"10.1016/j.asoc.2024.112501","DOIUrl":"10.1016/j.asoc.2024.112501","url":null,"abstract":"<div><div>The integration of Additive Manufacturing (AM) has the potential to transform Supply Chain (SC) dynamics, but its implementation introduces risks that require careful management. This paper presents an innovative optimization framework for evaluating AM integration within SCs through two policies aimed at creating resilience. By exploring the intersection of AM and Traditional Manufacturing (TM), the study focuses on restructuring SCs for full or partial product production using AM techniques. To enhance SC resilience against material shortages, smart contracts with buffer suppliers are employed. The main objective is to reduce operational and conventional costs while optimizing SC performance. To address cost uncertainty, this research introduces a novel Monte Carlo (MC) and Machine Learning (ML) hybrid approach, termed MCML. This method leverages MCML-Particle Swarm Optimization (MCML-PSO) and MCML-Genetic Algorithm (MCML-GA) for optimization. A real-world case study validates the model, showing that it reduces costs and improves the accuracy of cost uncertainty estimation compared to standalone TM and AM approaches. Various methods, including PSO, GA, MC-PSO, and MC-GA, were evaluated, with MCML-PSO demonstrating the best performance in minimizing total costs. This study highlights the benefits of integrating AM into SCs, emphasizing the importance of precise cost uncertainty estimation. The proposed model offers valuable insights for decision-makers, helping them design resilient and efficient SCs while mitigating the risks associated with AM technology.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112501"},"PeriodicalIF":7.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745495","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":"GRASP with Evolutionary Path Relinking for the conditional p-Dispersion problem","authors":"Jesús Sánchez-Oro ,&nbsp;Anna Martínez-Gavara ,&nbsp;Ana D. López-Sánchez ,&nbsp;Rafael Martí ,&nbsp;Abraham Duarte","doi":"10.1016/j.asoc.2024.112494","DOIUrl":"10.1016/j.asoc.2024.112494","url":null,"abstract":"<div><div>In this paper, we propose a new heuristic method that hybridizes GRASP with Path Relinking to solve the conditional <span><math><mi>p</mi></math></span>-Dispersion problem. Given <span><math><mi>n</mi></math></span> elements, from which <span><math><mrow><mi>q</mi><mo>&lt;</mo><mi>n</mi></mrow></math></span> have been already selected, this problem seeks to select <span><math><mrow><mi>p</mi><mo>&lt;</mo><mi>n</mi></mrow></math></span> additional unselected elements to maximize the minimum dissimilarity among them. The conditional <span><math><mi>p</mi></math></span>-dispersion problem models a facility location problem motivated by a real situation faced in many practical settings arising when some facilities have been already located. The algorithm includes a novel proposal based on an efficient interplay between search intensification and diversification provided by the Path Relinking component, and it also incorporates an intelligent way to measure the diversity among solutions. An extensive computational experimentation is carried out to compare the performance of our heuristic with the state of the art method. The comparison shows that our proposal is competitive with the existing method, since it is able to identify 17 best-known values. Additionally, our experimentation includes a real practical case solved for a Spanish company in its expansion process. This case illustrates both the applicability of the conditional <span><math><mi>p</mi></math></span>-dispersion model, and the suitability of our algorithm to efficiently solve practical instances.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112494"},"PeriodicalIF":7.2,"publicationDate":"2024-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745496","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}

{"title":"Adolescent mental health state assessment framework by combining YOLO with random forest","authors":"Min Wan ,&nbsp;Sai Zou","doi":"10.1016/j.asoc.2024.112497","DOIUrl":"10.1016/j.asoc.2024.112497","url":null,"abstract":"<div><div>The problems of adolescent mental health are becoming increasingly prominent. The house-tree-person test (HTPT) method can map the psychological state of subjects and has been widely used in clinical testing. However, the HTPT method requires an amount of time for professionals to assess. Based on the HTPT method, how to use artificial intelligence technology to quickly, objectively, and automatically complete mental state assessments has become a new trend. In this paper, a Hybrid of Enhanced YOLO and Random Forest algorithm for adolescent mental health assessment is proposed. Because of the dependence between HTPT feature positions, Bayesian theory is used to enhance YOLO to improve detection accuracy. Among the many features detected by the enhanced YOLO algorithm, RF is used to automatically assess mental state. The method is validated by simulation experiments and actual measurements of university students. Moreover, the simulation results show that the recognition accuracy can reach 92% and the recognition speed can reach the second level. The measured results show that this method can quickly and accurately assess mental state.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112497"},"PeriodicalIF":7.2,"publicationDate":"2024-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723389","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":"Leveraging more information for blind stereo super-resolution via large-window cross-attention","authors":"Weifeng Cao,&nbsp;Xiaoyan Lei,&nbsp;Yong Jiang,&nbsp;Zongfei Bai,&nbsp;Xiaoliang Qian","doi":"10.1016/j.asoc.2024.112492","DOIUrl":"10.1016/j.asoc.2024.112492","url":null,"abstract":"<div><div>Stereo image super-resolution aims to reconstruct high-resolution images by effectively utilizing cross-view complementary information from stereo image pairs. A prevalent method in Stereo image super-resolution is the stereo cross-attention module, which allows the model to focus on and integrate relevant features from both the left and right views. Despite its advantages, our analysis using a diagnostic tool called local attribution map (LAM) reveals that current methods exhibit limitations in effectively leveraging this complementary information. To address this issue, we propose the Double Stereo Cross-Attention Module (DSCAM), which utilizes an Overlapping Stereo Cross-Attention (OSCA) mechanism that enhances the integration of cross-view complementary information by using overlapping windows, followed by an additional multiplication step to refine and emphasize the combined features. Additionally, we develop a stereo image degradation model that ensures the consistency of degradation between stereo pairs, accurately simulating the real-world degradation process of stereo images. Extensive experiments have demonstrated that our method achieves visually pleasing results, making it the first to address the problem of stereo image super-resolution in real-world scenarios. The source code is available at <span><span>https://github.com/nathan66666/LCASSR.git</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112492"},"PeriodicalIF":7.2,"publicationDate":"2024-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142723391","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":"Viable intertwined supply network: Modelling and dynamic analysis using artificial neural networks","authors":"Shahid Ahmad Bhat ,&nbsp;Tariq Aljuneidi","doi":"10.1016/j.asoc.2024.112503","DOIUrl":"10.1016/j.asoc.2024.112503","url":null,"abstract":"<div><div>The viability of intertwined supply networks (ISNs) has recently been studied as a critical topic in operations management. Modelling the viability of ISNs is considered a promising tool to meet the demands of extraordinary events, such as the Russo-Ukrainian War and the COVID-19 pandemic. To enhance the viability of ISNs, the structures of ISN must be modelled, and the behavioral dynamics of interactions between firms in a changing environment should be analyzed. In this study, a trophic chain-based dynamic formulation of ISN viability is presented, and a solution methodology for dynamic analysis of the ISN viability model is designed. The concept of artificial neural networks (ANNs) in ISN analysis is introduced to predict and analyze future behavioral strategies in buyer-supplier relations. The dynamic model of ISN is represented by a system of nonlinear differential equations and described in terms of three dynamic values: suppliers <span><math><mrow><mi>X</mi><mrow><mfenced><mrow><mi>τ</mi></mrow></mfenced></mrow></mrow></math></span>, focal firms <span><math><mrow><mi>Y</mi><mrow><mfenced><mrow><mi>τ</mi></mrow></mfenced></mrow></mrow></math></span>, and market demand <span><math><mrow><mi>Z</mi><mrow><mfenced><mrow><mi>τ</mi></mrow></mfenced></mrow></mrow></math></span>. The stochastic numerical simulations are performed for the dynamics of ISN model by employing ANNs with a scaled conjugate gradient neural network (SCGNN) in a more advanced and efficient manner. Two numerical cases are investigated to evaluate the performance of the proposed approach. The validation, correctness, and reliability of the proposed stochastic SCGNN technique are analyzed by selecting 78 % of the data for training, 12 % for validation, and 10 % for testing. The correctness of the scheme is authenticated through the overlapping of the proposed and state-of-the-art results. Moreover, the statistical analysis is presented graphically in terms of mean square error, state transitions, function fitness, error histograms. The regression coefficient values are calculated as 1 for each scenario presents the perfect model. Finally, a comparison of numerical results, which shows the overlapping is examined and the absolute error is performed between 10⁻⁰⁵ to 10⁻⁰⁷. The small mean square error performances enhance the correctness of the scheme. These results indicate that the dynamical model can effectively analyze the ISN structures and help researchers and practitioners ensure the survival of supply chains during extraordinary events.</div></div>","PeriodicalId":50737,"journal":{"name":"Applied Soft Computing","volume":"168 ","pages":"Article 112503"},"PeriodicalIF":7.2,"publicationDate":"2024-11-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142745493","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}