Hydrogen partitioning between liquid iron alloys and silicate melts governs its distribution and cycling in Earth’s deep interior.Existing models based on simplified Fe-H systems predict strong hydrogen sequestration...Hydrogen partitioning between liquid iron alloys and silicate melts governs its distribution and cycling in Earth’s deep interior.Existing models based on simplified Fe-H systems predict strong hydrogen sequestration into the core.However,these models do not account for the modulating effects of major light elements such as oxygen and silicon in the core during Earth’s primordial differentiation.In this study,we use first-principles molecular dynamics simulations,augmented by machine learning techniques,to quantify hydrogen chemical potentials in quaternary Fe-O-Si-H systems under early core-mantle boundary conditions(135 GPa,5000 K).Our results demonstrate that the presence of 5.2 wt%oxygen and 4.8 wt%silicon reduces the siderophile affinity of hydrogen by 35%,decreasing its alloy-silicate partition coefficient from 18.2(in the case of Fe-H)to 11.8(in the case of Fe-O-Si-H).These findings suggest that previous estimates of the core hydrogen content derived from binary system models require downward revision.Our study underscores the critical role of multicomponent interactions in core formation models and provides first-principles-derived constraints to reconcile Earth’s present-day hydrogen reservoirs with its accretionary history.展开更多
Deep Learning-based systems for Finger vein recognition have gained rising attention in recent years due to improved efficiency and enhanced security.The performance of existing CNN-based methods is limited by the pun...Deep Learning-based systems for Finger vein recognition have gained rising attention in recent years due to improved efficiency and enhanced security.The performance of existing CNN-based methods is limited by the puny generalization of learned features and deficiency of the finger vein image training data.Considering the concerns of existing methods,in this work,a simplified deep transfer learning-based framework for finger-vein recognition is developed using an EfficientNet model of deep learning with a self-attention mechanism.Data augmentation using various geometrical methods is employed to address the problem of training data shortage required for a deep learning model.The proposed model is tested using K-fold cross-validation on three publicly available datasets:HKPU,FVUSM,and SDUMLA.Also,the developed network is compared with other modern deep nets to check its effectiveness.In addition,a comparison of the proposed method with other existing Finger vein recognition(FVR)methods is also done.The experimental results exhibited superior recognition accuracy of the proposed method compared to other existing methods.In addition,the developed method proves to be more effective and less sophisticated at extracting robust features.The proposed EffAttenNet achieves an accuracy of 98.14%on HKPU,99.03%on FVUSM,and 99.50%on SDUMLA databases.展开更多
The article employs the wetlands of Ruoergai(i.e.,Zoige),Sichuan Province,as a case study to analyze changes over various time scales,utilizing Landsat data from 2004,2008,2012,2016,2020,and 2023.The study uses the GE...The article employs the wetlands of Ruoergai(i.e.,Zoige),Sichuan Province,as a case study to analyze changes over various time scales,utilizing Landsat data from 2004,2008,2012,2016,2020,and 2023.The study uses the GEE platform and a deep learning model,focusing on the long-term perspective.This analysis serves as a focal point for discussing sustainable development,offering ecological balance information and a realistic foundation.The paper systematically gathers remote sensing classification images resembling sample points on the GEE(Google Earth Engine)platform.Simultaneously,it develops a deep learning model for classifying land types in Ruoergai into six categories:river-wetland,lake-wetland,swamp-wetland,grassland,forest and shrubland.This classification is achieved by utilizing various bands of Landsat data as input features and assigning land cover as corresponding labels.A comparison of classification results in 2016 indicates that the approach integrating the GEE platform and the deep learning model enhances overall accuracy by 9%compared to the random forest method.Furthermore,the overall accuracy surpasses that of the support vector machine method by 16%,and the CART method by 23%.These results affirm that the combined GEE platform and deep learning model outperforms the random forest method in overall accuracy.The findings reveal a declining trend in the wetland area of Ruoergai from 2004 to 2012,with the area remaining relatively stable from 2012 to 2016.Subsequently,there is a significant increase from 2016 to 2023.These trends corroborate the positive outcomes of long-term environmental protection policies implemented by the Chinese government.Furthermore,they underscore the success and efforts exerted by both the government and society in the sustainable management of wetland ecosystems.This serves as an exemplary case for advancing the SDG 15.1 development goal.展开更多
The rise of deep learning has brought about transformative advancements in both scientific research and engineering applications.The 2024 Nobel Prizes,particularly in Physics and Chemistry,highlighted the revolutionar...The rise of deep learning has brought about transformative advancements in both scientific research and engineering applications.The 2024 Nobel Prizes,particularly in Physics and Chemistry,highlighted the revolutionary impact of deep learning,with AlphaFold’s breakthrough in protein structure prediction exemplifying its potential.This review explores the historical evolution of deep learning,from its foundational theories in neural networks and connectionism to its modern applications in various fields.Focus is given to its use in geotechnical engineering,particularly in geological disaster prediction,tunnel safety monitoring,and structural design optimization.The integration of deep learning models such as Convolutional Neural Networks(CNNs),Recurrent Neural Networks(RNNs),and Transformers has enabled significant progress in analyzing complex,unstructured data,offering innovative solutions to longstanding engineering challenges.The review also examines the opportunities and challenges faced by the field,advocating for interdisciplinary collaboration and open data sharing to further unlock deep learning’s potential in advancing both scientific and engineering disciplines.As deep learning continues to evolve,it promises to drive further innovation,shaping the future of engineering practices and scientific discovery.展开更多
The development of digital twins for geotechnical structures necessitates the real-time updates of threedimensional(3D)virtual models(e.g.numerical finite element method(FEM)model)to accurately predict time-varying ge...The development of digital twins for geotechnical structures necessitates the real-time updates of threedimensional(3D)virtual models(e.g.numerical finite element method(FEM)model)to accurately predict time-varying geotechnical responses(e.g.consolidation settlement)in a 3D spatial domain.However,traditional 3D numerical model updating approaches are computationally prohibitive and therefore difficult to update the 3D responses in real time.To address these challenges,this study proposes a novel machine learning framework called sparse dictionary learning(T-3D-SDL)for real-time updating of time-varying 3D geotechnical responses.In T-3D-SDL,a concerned dataset(e.g.time-varying 3D settlement)is approximated as a linear superposition of dictionary atoms generated from 3D random FEM analyses.Field monitoring data are then used to identify non-trivial atoms and estimate their weights within a Bayesian framework for model updating and prediction.The proposed approach enables the real-time update of temporally varying settlements with a high 3D spatial resolution and quantified uncertainty as field monitoring data evolve.The proposed approach is illustrated using an embankment construction project.The results show that the proposed approach effectively improves settlement predictions along temporal and 3D spatial dimensions,with minimal latency(e.g.within minutes),as monitoring data appear.In addition,the proposed approach requires only a reasonably small number of 3D FEM model evaluations,avoids the use of widely adopted yet often criticized surrogate models,and effectively addresses the limitations(e.g.computational inefficiency)of existing 3D model updating approaches.展开更多
为解决现有辨识方法在针对耦合的次/超同步振荡参数提取过程中的噪声适应性差和模态混叠问题,该文提出了一种自适应的变分模态分解法(variational mode decomposition,VMD),定义残差损失总熵、中心频率的切比雪夫距离以及边缘熵共同决...为解决现有辨识方法在针对耦合的次/超同步振荡参数提取过程中的噪声适应性差和模态混叠问题,该文提出了一种自适应的变分模态分解法(variational mode decomposition,VMD),定义残差损失总熵、中心频率的切比雪夫距离以及边缘熵共同决定分解模态数和带宽,结合最小二乘-旋转不变技术(total least square-estimating signal parameter via rotational invariance techniques,TLS-ESPRIT)对分解出的振荡分量进行参数辨识,无需另外使用降噪算法。通过复合信号测试法、PSCAD/EMTDC电磁暂态仿真法验证了所提方法的有效性。最后,将所提方法与改进Prony算法、MCEEMD法在不同噪声水平和振荡频率下进行对比,结果表明,所提方法能够有效地抑制原始信号的噪声干扰,对耦合的次/超同步振荡信号分解更加准确,参数辨识结果可靠性较高,对风电系统振荡溯源、改善系统阻尼具有一定的参考意义。展开更多
Early identification and treatment of stroke can greatly improve patient outcomes and quality of life.Although clinical tests such as the Cincinnati Pre-hospital Stroke Scale(CPSS)and the Face Arm Speech Test(FAST)are...Early identification and treatment of stroke can greatly improve patient outcomes and quality of life.Although clinical tests such as the Cincinnati Pre-hospital Stroke Scale(CPSS)and the Face Arm Speech Test(FAST)are commonly used for stroke screening,accurate administration is dependent on specialized training.In this study,we proposed a novel multimodal deep learning approach,based on the FAST,for assessing suspected stroke patients exhibiting symptoms such as limb weakness,facial paresis,and speech disorders in acute settings.We collected a dataset comprising videos and audio recordings of emergency room patients performing designated limb movements,facial expressions,and speech tests based on the FAST.We compared the constructed deep learning model,which was designed to process multi-modal datasets,with six prior models that achieved good action classification performance,including the I3D,SlowFast,X3D,TPN,TimeSformer,and MViT.We found that the findings of our deep learning model had a higher clinical value compared with the other approaches.Moreover,the multi-modal model outperformed its single-module variants,highlighting the benefit of utilizing multiple types of patient data,such as action videos and speech audio.These results indicate that a multi-modal deep learning model combined with the FAST could greatly improve the accuracy and sensitivity of early stroke identification of stroke,thus providing a practical and powerful tool for assessing stroke patients in an emergency clinical setting.展开更多
With the rapid development of artificial intelligence,the Internet of Things(IoT)can deploy various machine learning algorithms for network and application management.In the IoT environment,many sensors and devices ge...With the rapid development of artificial intelligence,the Internet of Things(IoT)can deploy various machine learning algorithms for network and application management.In the IoT environment,many sensors and devices generatemassive data,but data security and privacy protection have become a serious challenge.Federated learning(FL)can achieve many intelligent IoT applications by training models on local devices and allowing AI training on distributed IoT devices without data sharing.This review aims to deeply explore the combination of FL and the IoT,and analyze the application of federated learning in the IoT from the aspects of security and privacy protection.In this paper,we first describe the potential advantages of FL and the challenges faced by current IoT systems in the fields of network burden and privacy security.Next,we focus on exploring and analyzing the advantages of the combination of FL on the Internet,including privacy security,attack detection,efficient communication of the IoT,and enhanced learning quality.We also list various application scenarios of FL on the IoT.Finally,we propose several open research challenges and possible solutions.展开更多
Ransomware has emerged as a critical cybersecurity threat,characterized by its ability to encrypt user data or lock devices,demanding ransom for their release.Traditional ransomware detection methods face limitations ...Ransomware has emerged as a critical cybersecurity threat,characterized by its ability to encrypt user data or lock devices,demanding ransom for their release.Traditional ransomware detection methods face limitations due to their assumption of similar data distributions between training and testing phases,rendering them less effective against evolving ransomware families.This paper introduces TLERAD(Transfer Learning for Enhanced Ransomware Attack Detection),a novel approach that leverages unsupervised transfer learning and co-clustering techniques to bridge the gap between source and target domains,enabling robust detection of both known and unknown ransomware variants.The proposed method achieves high detection accuracy,with an AUC of 0.98 for known ransomware and 0.93 for unknown ransomware,significantly outperforming baseline methods.Comprehensive experiments demonstrate TLERAD’s effectiveness in real-world scenarios,highlighting its adapt-ability to the rapidly evolving ransomware landscape.The paper also discusses future directions for enhancing TLERAD,including real-time adaptation,integration with lightweight and post-quantum cryptography,and the incorporation of explainable AI techniques.展开更多
In the realm of Intelligent Railway Transportation Systems,effective multi-party collaboration is crucial due to concerns over privacy and data silos.Vertical Federated Learning(VFL)has emerged as a promising approach...In the realm of Intelligent Railway Transportation Systems,effective multi-party collaboration is crucial due to concerns over privacy and data silos.Vertical Federated Learning(VFL)has emerged as a promising approach to facilitate such collaboration,allowing diverse entities to collectively enhance machine learning models without the need to share sensitive training data.However,existing works have highlighted VFL’s susceptibility to privacy inference attacks,where an honest but curious server could potentially reconstruct a client’s raw data from embeddings uploaded by the client.This vulnerability poses a significant threat to VFL-based intelligent railway transportation systems.In this paper,we introduce SensFL,a novel privacy-enhancing method to against privacy inference attacks in VFL.Specifically,SensFL integrates regularization of the sensitivity of embeddings to the original data into the model training process,effectively limiting the information contained in shared embeddings.By reducing the sensitivity of embeddings to the original data,SensFL can effectively resist reverse privacy attacks and prevent the reconstruction of the original data from the embeddings.Extensive experiments were conducted on four distinct datasets and three different models to demonstrate the efficacy of SensFL.Experiment results show that SensFL can effectively mitigate privacy inference attacks while maintaining the accuracy of the primary learning task.These results underscore SensFL’s potential to advance privacy protection technologies within VFL-based intelligent railway systems,addressing critical security concerns in collaborative learning environments.展开更多
基金supported by the National Key R&D Program of China(Grant No.2022YFF0503203)National Natural Science Foundation of China(NSFC)projects(Grant Nos.42441826 and 42173041)+1 种基金the Key Research Program of the Institute of Geology and Geophysics,Chinese Academy of Sciences(Grant No.IGGCAS-202204)the computational facilities of the Computer Simulation Laboratory at IGGCAS and the Beijing Super Cloud Computing Center(BSCC).
文摘Hydrogen partitioning between liquid iron alloys and silicate melts governs its distribution and cycling in Earth’s deep interior.Existing models based on simplified Fe-H systems predict strong hydrogen sequestration into the core.However,these models do not account for the modulating effects of major light elements such as oxygen and silicon in the core during Earth’s primordial differentiation.In this study,we use first-principles molecular dynamics simulations,augmented by machine learning techniques,to quantify hydrogen chemical potentials in quaternary Fe-O-Si-H systems under early core-mantle boundary conditions(135 GPa,5000 K).Our results demonstrate that the presence of 5.2 wt%oxygen and 4.8 wt%silicon reduces the siderophile affinity of hydrogen by 35%,decreasing its alloy-silicate partition coefficient from 18.2(in the case of Fe-H)to 11.8(in the case of Fe-O-Si-H).These findings suggest that previous estimates of the core hydrogen content derived from binary system models require downward revision.Our study underscores the critical role of multicomponent interactions in core formation models and provides first-principles-derived constraints to reconcile Earth’s present-day hydrogen reservoirs with its accretionary history.
文摘Deep Learning-based systems for Finger vein recognition have gained rising attention in recent years due to improved efficiency and enhanced security.The performance of existing CNN-based methods is limited by the puny generalization of learned features and deficiency of the finger vein image training data.Considering the concerns of existing methods,in this work,a simplified deep transfer learning-based framework for finger-vein recognition is developed using an EfficientNet model of deep learning with a self-attention mechanism.Data augmentation using various geometrical methods is employed to address the problem of training data shortage required for a deep learning model.The proposed model is tested using K-fold cross-validation on three publicly available datasets:HKPU,FVUSM,and SDUMLA.Also,the developed network is compared with other modern deep nets to check its effectiveness.In addition,a comparison of the proposed method with other existing Finger vein recognition(FVR)methods is also done.The experimental results exhibited superior recognition accuracy of the proposed method compared to other existing methods.In addition,the developed method proves to be more effective and less sophisticated at extracting robust features.The proposed EffAttenNet achieves an accuracy of 98.14%on HKPU,99.03%on FVUSM,and 99.50%on SDUMLA databases.
文摘The article employs the wetlands of Ruoergai(i.e.,Zoige),Sichuan Province,as a case study to analyze changes over various time scales,utilizing Landsat data from 2004,2008,2012,2016,2020,and 2023.The study uses the GEE platform and a deep learning model,focusing on the long-term perspective.This analysis serves as a focal point for discussing sustainable development,offering ecological balance information and a realistic foundation.The paper systematically gathers remote sensing classification images resembling sample points on the GEE(Google Earth Engine)platform.Simultaneously,it develops a deep learning model for classifying land types in Ruoergai into six categories:river-wetland,lake-wetland,swamp-wetland,grassland,forest and shrubland.This classification is achieved by utilizing various bands of Landsat data as input features and assigning land cover as corresponding labels.A comparison of classification results in 2016 indicates that the approach integrating the GEE platform and the deep learning model enhances overall accuracy by 9%compared to the random forest method.Furthermore,the overall accuracy surpasses that of the support vector machine method by 16%,and the CART method by 23%.These results affirm that the combined GEE platform and deep learning model outperforms the random forest method in overall accuracy.The findings reveal a declining trend in the wetland area of Ruoergai from 2004 to 2012,with the area remaining relatively stable from 2012 to 2016.Subsequently,there is a significant increase from 2016 to 2023.These trends corroborate the positive outcomes of long-term environmental protection policies implemented by the Chinese government.Furthermore,they underscore the success and efforts exerted by both the government and society in the sustainable management of wetland ecosystems.This serves as an exemplary case for advancing the SDG 15.1 development goal.
基金support provided by the Hebei Province Full-Time Recruitment of National High-Level Innovative Talents Research Project(Grant No.2023HBQZYCSB004).
文摘The rise of deep learning has brought about transformative advancements in both scientific research and engineering applications.The 2024 Nobel Prizes,particularly in Physics and Chemistry,highlighted the revolutionary impact of deep learning,with AlphaFold’s breakthrough in protein structure prediction exemplifying its potential.This review explores the historical evolution of deep learning,from its foundational theories in neural networks and connectionism to its modern applications in various fields.Focus is given to its use in geotechnical engineering,particularly in geological disaster prediction,tunnel safety monitoring,and structural design optimization.The integration of deep learning models such as Convolutional Neural Networks(CNNs),Recurrent Neural Networks(RNNs),and Transformers has enabled significant progress in analyzing complex,unstructured data,offering innovative solutions to longstanding engineering challenges.The review also examines the opportunities and challenges faced by the field,advocating for interdisciplinary collaboration and open data sharing to further unlock deep learning’s potential in advancing both scientific and engineering disciplines.As deep learning continues to evolve,it promises to drive further innovation,shaping the future of engineering practices and scientific discovery.
基金supported by a grant from the Research Grant Council of Hong Kong Special Administrative Region(Project No.11207724).
文摘The development of digital twins for geotechnical structures necessitates the real-time updates of threedimensional(3D)virtual models(e.g.numerical finite element method(FEM)model)to accurately predict time-varying geotechnical responses(e.g.consolidation settlement)in a 3D spatial domain.However,traditional 3D numerical model updating approaches are computationally prohibitive and therefore difficult to update the 3D responses in real time.To address these challenges,this study proposes a novel machine learning framework called sparse dictionary learning(T-3D-SDL)for real-time updating of time-varying 3D geotechnical responses.In T-3D-SDL,a concerned dataset(e.g.time-varying 3D settlement)is approximated as a linear superposition of dictionary atoms generated from 3D random FEM analyses.Field monitoring data are then used to identify non-trivial atoms and estimate their weights within a Bayesian framework for model updating and prediction.The proposed approach enables the real-time update of temporally varying settlements with a high 3D spatial resolution and quantified uncertainty as field monitoring data evolve.The proposed approach is illustrated using an embankment construction project.The results show that the proposed approach effectively improves settlement predictions along temporal and 3D spatial dimensions,with minimal latency(e.g.within minutes),as monitoring data appear.In addition,the proposed approach requires only a reasonably small number of 3D FEM model evaluations,avoids the use of widely adopted yet often criticized surrogate models,and effectively addresses the limitations(e.g.computational inefficiency)of existing 3D model updating approaches.
文摘为解决现有辨识方法在针对耦合的次/超同步振荡参数提取过程中的噪声适应性差和模态混叠问题,该文提出了一种自适应的变分模态分解法(variational mode decomposition,VMD),定义残差损失总熵、中心频率的切比雪夫距离以及边缘熵共同决定分解模态数和带宽,结合最小二乘-旋转不变技术(total least square-estimating signal parameter via rotational invariance techniques,TLS-ESPRIT)对分解出的振荡分量进行参数辨识,无需另外使用降噪算法。通过复合信号测试法、PSCAD/EMTDC电磁暂态仿真法验证了所提方法的有效性。最后,将所提方法与改进Prony算法、MCEEMD法在不同噪声水平和振荡频率下进行对比,结果表明,所提方法能够有效地抑制原始信号的噪声干扰,对耦合的次/超同步振荡信号分解更加准确,参数辨识结果可靠性较高,对风电系统振荡溯源、改善系统阻尼具有一定的参考意义。
基金supported by the Ministry of Science and Technology of China,No.2020AAA0109605(to XL)Meizhou Major Scientific and Technological Innovation PlatformsProjects of Guangdong Provincial Science & Technology Plan Projects,No.2019A0102005(to HW).
文摘Early identification and treatment of stroke can greatly improve patient outcomes and quality of life.Although clinical tests such as the Cincinnati Pre-hospital Stroke Scale(CPSS)and the Face Arm Speech Test(FAST)are commonly used for stroke screening,accurate administration is dependent on specialized training.In this study,we proposed a novel multimodal deep learning approach,based on the FAST,for assessing suspected stroke patients exhibiting symptoms such as limb weakness,facial paresis,and speech disorders in acute settings.We collected a dataset comprising videos and audio recordings of emergency room patients performing designated limb movements,facial expressions,and speech tests based on the FAST.We compared the constructed deep learning model,which was designed to process multi-modal datasets,with six prior models that achieved good action classification performance,including the I3D,SlowFast,X3D,TPN,TimeSformer,and MViT.We found that the findings of our deep learning model had a higher clinical value compared with the other approaches.Moreover,the multi-modal model outperformed its single-module variants,highlighting the benefit of utilizing multiple types of patient data,such as action videos and speech audio.These results indicate that a multi-modal deep learning model combined with the FAST could greatly improve the accuracy and sensitivity of early stroke identification of stroke,thus providing a practical and powerful tool for assessing stroke patients in an emergency clinical setting.
基金supported by the Shandong Province Science and Technology Project(2023TSGC0509,2022TSGC2234)Qingdao Science and Technology Plan Project(23-1-5-yqpy-2-qy)Open Topic Grants of Anhui Province Key Laboratory of Intelligent Building&Building Energy Saving,Anhui Jianzhu University(IBES2024KF08).
文摘With the rapid development of artificial intelligence,the Internet of Things(IoT)can deploy various machine learning algorithms for network and application management.In the IoT environment,many sensors and devices generatemassive data,but data security and privacy protection have become a serious challenge.Federated learning(FL)can achieve many intelligent IoT applications by training models on local devices and allowing AI training on distributed IoT devices without data sharing.This review aims to deeply explore the combination of FL and the IoT,and analyze the application of federated learning in the IoT from the aspects of security and privacy protection.In this paper,we first describe the potential advantages of FL and the challenges faced by current IoT systems in the fields of network burden and privacy security.Next,we focus on exploring and analyzing the advantages of the combination of FL on the Internet,including privacy security,attack detection,efficient communication of the IoT,and enhanced learning quality.We also list various application scenarios of FL on the IoT.Finally,we propose several open research challenges and possible solutions.
文摘Ransomware has emerged as a critical cybersecurity threat,characterized by its ability to encrypt user data or lock devices,demanding ransom for their release.Traditional ransomware detection methods face limitations due to their assumption of similar data distributions between training and testing phases,rendering them less effective against evolving ransomware families.This paper introduces TLERAD(Transfer Learning for Enhanced Ransomware Attack Detection),a novel approach that leverages unsupervised transfer learning and co-clustering techniques to bridge the gap between source and target domains,enabling robust detection of both known and unknown ransomware variants.The proposed method achieves high detection accuracy,with an AUC of 0.98 for known ransomware and 0.93 for unknown ransomware,significantly outperforming baseline methods.Comprehensive experiments demonstrate TLERAD’s effectiveness in real-world scenarios,highlighting its adapt-ability to the rapidly evolving ransomware landscape.The paper also discusses future directions for enhancing TLERAD,including real-time adaptation,integration with lightweight and post-quantum cryptography,and the incorporation of explainable AI techniques.
基金supported by Systematic Major Project of Shuohuang Railway Development Co.,Ltd.,National Energy Group(Grant Number:SHTL-23-31)Beijing Natural Science Foundation(U22B2027).
文摘In the realm of Intelligent Railway Transportation Systems,effective multi-party collaboration is crucial due to concerns over privacy and data silos.Vertical Federated Learning(VFL)has emerged as a promising approach to facilitate such collaboration,allowing diverse entities to collectively enhance machine learning models without the need to share sensitive training data.However,existing works have highlighted VFL’s susceptibility to privacy inference attacks,where an honest but curious server could potentially reconstruct a client’s raw data from embeddings uploaded by the client.This vulnerability poses a significant threat to VFL-based intelligent railway transportation systems.In this paper,we introduce SensFL,a novel privacy-enhancing method to against privacy inference attacks in VFL.Specifically,SensFL integrates regularization of the sensitivity of embeddings to the original data into the model training process,effectively limiting the information contained in shared embeddings.By reducing the sensitivity of embeddings to the original data,SensFL can effectively resist reverse privacy attacks and prevent the reconstruction of the original data from the embeddings.Extensive experiments were conducted on four distinct datasets and three different models to demonstrate the efficacy of SensFL.Experiment results show that SensFL can effectively mitigate privacy inference attacks while maintaining the accuracy of the primary learning task.These results underscore SensFL’s potential to advance privacy protection technologies within VFL-based intelligent railway systems,addressing critical security concerns in collaborative learning environments.
