In image analysis,high-precision semantic segmentation predominantly relies on supervised learning.Despite significant advancements driven by deep learning techniques,challenges such as class imbalance and dynamic per...In image analysis,high-precision semantic segmentation predominantly relies on supervised learning.Despite significant advancements driven by deep learning techniques,challenges such as class imbalance and dynamic performance evaluation persist.Traditional weighting methods,often based on pre-statistical class counting,tend to overemphasize certain classes while neglecting others,particularly rare sample categories.Approaches like focal loss and other rare-sample segmentation techniques introduce multiple hyperparameters that require manual tuning,leading to increased experimental costs due to their instability.This paper proposes a novel CAWASeg framework to address these limitations.Our approach leverages Grad-CAM technology to generate class activation maps,identifying key feature regions that the model focuses on during decision-making.We introduce a Comprehensive Segmentation Performance Score(CSPS)to dynamically evaluate model performance by converting these activation maps into pseudo mask and comparing them with Ground Truth.Additionally,we design two adaptive weights for each class:a Basic Weight(BW)and a Ratio Weight(RW),which the model adjusts during training based on real-time feedback.Extensive experiments on the COCO-Stuff,CityScapes,and ADE20k datasets demonstrate that our CAWASeg framework significantly improves segmentation performance for rare sample categories while enhancing overall segmentation accuracy.The proposed method offers a robust and efficient solution for addressing class imbalance in semantic segmentation tasks.展开更多
The photovoltaic(PV)output process is inherently complex,often disrupted by a multitude of meteoro-logical factors,while conventional detection methods at PV power stations prove inadequate,compromising prediction acc...The photovoltaic(PV)output process is inherently complex,often disrupted by a multitude of meteoro-logical factors,while conventional detection methods at PV power stations prove inadequate,compromising prediction accuracy.To address this challenge,this paper introduces a power prediction method that leverages modal switching(MS),weight factor adjustment(WFA),and parallel long short-term memory(PALSTM).Initially,historical PV power station data is categorized into distinct modes based on global horizontal irradiance and converted solar angles.Correlation analysis is then employed to evaluate the impact of various meteorological factors on PV power,selecting those with strong correlations for each specific mode.Subsequently,the weights of meteorological parameters are optimized and adjusted,and a PALSTM neural network is constructed,with its parallel modal parameters refined through training.Depending on the prediction time and input data mode characteristics,the appropriate mode channel is selected to forecast PV power station generation.Ultimately,the feasibility of this method is validated through an illustrative analysis of measured data from an Australian PV power station.Comparative test results underscore the method’s advantages,particularly in scenarios where existing detection methods are lacking and meteorological factors frequently fluctuate,demonstrating its superior prediction accuracy and stability.展开更多
A hierarchical scheme of feature-based model similarity measurement was proposed,named CSG_D2,in which both geometry similarity and topology similarity were applied.The features of 3D mechanical part were constructed ...A hierarchical scheme of feature-based model similarity measurement was proposed,named CSG_D2,in which both geometry similarity and topology similarity were applied.The features of 3D mechanical part were constructed by a series of primitive features with tree structure,as a form of constructive solid geometry(CSG) tree.The D2 shape distributions of these features were extracted for geometry similarity measurement,and the pose vector and non-disappeared proportion of each leaf node were gained for topology similarity measurement.Based on these,the dissimilarity between the query and the candidate was accessed by level-by-level CSG tree comparisons.With the adjustable weights,our scheme satisfies different comparison emphasis on the geometry or topology similarity.The assessment results from CSG_D2 demonstrate more discriminative than those from D2 in the analysis of precision-recall and similarity matrix.Finally,an experimental search engine is applied for mechanical parts reuse by using CSG_D2,which is convenient for the mechanical design process.展开更多
The separate-layer injection in different interlayers and the injection of the same-molecular-weight polymer so- lution in a layer are necessary in the polymer flooding process because of heterogeneous multilayer sand...The separate-layer injection in different interlayers and the injection of the same-molecular-weight polymer so- lution in a layer are necessary in the polymer flooding process because of heterogeneous multilayer sandstone reservoirs in EOR projects. To alleviate the matching problems between the layer permeability and the injected polymer molecular weight, a molecular weight adjusting device with porous medium was designed on the basis of mechanical degradation principle. In terms of four variables （polymer concentration, pore diameter, length of shear component and flow rate ）, the theological behavior of hydrolyzed polyacrylamide （HPAM） solu- tion flowing through the device was investigated in detail. The change of these variables is able to control the shear rate of HPAM solutions through ceramic foam, and achieve the desired degree of shear degradation and the final theological parameters-viscosity loss, viscoelasticity and pressure drop. Therefore, a linear relationship between viscosity loss and shearing rate was established so as to obtain the targeted viscosity easily. Field tests in the Daqing Oil Field showed that the polymer molecular weight could drop 20% to 50%. In a word, the results could guide the industrial application of the novel device and the further study of polymer degradation flowing through the porous medium.展开更多
Fingerprint features,as unique and stable biometric identifiers,are crucial for identity verification.However,traditional centralized methods of processing these sensitive data linked to personal identity pose signifi...Fingerprint features,as unique and stable biometric identifiers,are crucial for identity verification.However,traditional centralized methods of processing these sensitive data linked to personal identity pose significant privacy risks,potentially leading to user data leakage.Federated Learning allows multiple clients to collaboratively train and optimize models without sharing raw data,effectively addressing privacy and security concerns.However,variations in fingerprint data due to factors such as region,ethnicity,sensor quality,and environmental conditions result in significant heterogeneity across clients.This heterogeneity adversely impacts the generalization ability of the global model,limiting its performance across diverse distributions.To address these challenges,we propose an Adaptive Federated Fingerprint Recognition algorithm(AFFR)based on Federated Learning.The algorithm incorporates a generalization adjustment mechanism that evaluates the generalization gap between the local models and the global model,adaptively adjusting aggregation weights to mitigate the impact of heterogeneity caused by differences in data quality and feature characteristics.Additionally,a noise mechanism is embedded in client-side training to reduce the risk of fingerprint data leakage arising from weight disclosures during model updates.Experiments conducted on three public datasets demonstrate that AFFR significantly enhances model accuracy while ensuring robust privacy protection,showcasing its strong application potential and competitiveness in heterogeneous data environments.展开更多
This paper proposes a particle swarm optimization（PSO） based particle filter（PF） tracking framework,the embedded PSO makes particles move toward the high likelihood area to find the optimal position in the state t...This paper proposes a particle swarm optimization（PSO） based particle filter（PF） tracking framework,the embedded PSO makes particles move toward the high likelihood area to find the optimal position in the state transition stage,and simultaneously incorporates the newest observations into the proposal distribution in the update stage.In the proposed approach,likelihood measure functions involving multiple features are presented to enhance the performance of model fitting.Furthermore,the multi-feature weights are self-adaptively adjusted by a PSO algorithm throughout the tracking process.There are three main contributions.Firstly,the PSO algorithm is fused into the PF framework,which can efficiently alleviate the particles degeneracy phenomenon.Secondly,an effective convergence criterion for the PSO algorithm is explored,which can avoid particles getting stuck in local minima and maintain a greater particle diversity.Finally,a multi-feature weight self-adjusting strategy is proposed,which can significantly improve the tracking robustness and accuracy.Experiments performed on several challenging public video sequences demonstrate that the proposed tracking approach achieves a considerable performance.展开更多
Nowadays, internet-based surveys are increasingly used for data collection, because their usage is simple and cheap. Also they give fast access to a large group of respondents. There are many factors affecting interne...Nowadays, internet-based surveys are increasingly used for data collection, because their usage is simple and cheap. Also they give fast access to a large group of respondents. There are many factors affecting internet surveys, such as measurement, survey design and sampling selection bias. The sampling has an important place in selection bias in internet survey. In terms of sample selection, the type of access to internet surveys has several limitations. There are internet surveys based on restricted access and on voluntary participation, and these are characterized by their implementation according to the type of survey. It can be used probability and non-probability sampling, both of which may lead to biased estimates. There are different ways to correct for selection biases;poststratification or weighting class adjustments, raking or rim weighting, generalized regression modeling and propensity score adjustments. This paper aims to describe methodological problems about selection bias issues and to give a review in internet surveys. Also the objective of this study is to show the effect of various correction techniques for reducing selection bias.展开更多
The multivariate linear errors-in-variables model when the regressors are missing at random in the sense of Rubin （1976） is considered in this paper. A constrained empirical likelihood confidence region for a parame...The multivariate linear errors-in-variables model when the regressors are missing at random in the sense of Rubin （1976） is considered in this paper. A constrained empirical likelihood confidence region for a parameter β0 in this model is proposed, which is constructed by combining the score function corresponding to the weighted squared orthogonal distance based on inverse probability with a constrained region of β0. It is shown that the empirical log-likelihood ratio at the true parameter converges to the standard chi-square distribution. Simulations show that the coverage rate of the proposed confidence region is closer to the nominal level and the length of confidence interval is narrower than those of the normal approximation of inverse probability weighted adjusted least square estimator in most cases. A real example is studied and the result supports the theory and simulation＇s conclusion.展开更多
In this paper, the parameters of a p-dimensional linear structural EV(error-in-variable)model are estimated when the coefficients vary with a real variable and the model error is time series.The adjust weighted least ...In this paper, the parameters of a p-dimensional linear structural EV(error-in-variable)model are estimated when the coefficients vary with a real variable and the model error is time series.The adjust weighted least squares(AWLS) method is used to estimate the parameters. It is shown that the estimators are weakly consistent and asymptotically normal, and the optimal convergence rate is also obtained. Simulations study are undertaken to illustrate our AWLSEs have good performance.展开更多
基金supported by the Funds for Central-Guided Local Science and Technology Development(Grant No.202407AC110005)Key Technologies for the Construction of a Whole-Process Intelligent Service System for Neuroendocrine Neoplasm.Supported by 2023 Opening Research Fund of Yunnan Key Laboratory of Digital Communications(YNJTKFB-20230686,YNKLDC-KFKT-202304).
文摘In image analysis,high-precision semantic segmentation predominantly relies on supervised learning.Despite significant advancements driven by deep learning techniques,challenges such as class imbalance and dynamic performance evaluation persist.Traditional weighting methods,often based on pre-statistical class counting,tend to overemphasize certain classes while neglecting others,particularly rare sample categories.Approaches like focal loss and other rare-sample segmentation techniques introduce multiple hyperparameters that require manual tuning,leading to increased experimental costs due to their instability.This paper proposes a novel CAWASeg framework to address these limitations.Our approach leverages Grad-CAM technology to generate class activation maps,identifying key feature regions that the model focuses on during decision-making.We introduce a Comprehensive Segmentation Performance Score(CSPS)to dynamically evaluate model performance by converting these activation maps into pseudo mask and comparing them with Ground Truth.Additionally,we design two adaptive weights for each class:a Basic Weight(BW)and a Ratio Weight(RW),which the model adjusts during training based on real-time feedback.Extensive experiments on the COCO-Stuff,CityScapes,and ADE20k datasets demonstrate that our CAWASeg framework significantly improves segmentation performance for rare sample categories while enhancing overall segmentation accuracy.The proposed method offers a robust and efficient solution for addressing class imbalance in semantic segmentation tasks.
基金This work was supported in part by the Natural Science Foundation of Henan Province,and the specific grant number is 232300420301the initial of author is P.L.,the URL to the sponsors’websites is https://kjt.henan.gov.cn/(accessed on 09 February 2025)And this work was also supported in part by the Fundamental Research Funds for the Universities of Henan Province,and the specific grant number is NSFRF220425,the initial of author is P.L.,the URL to sponsors websites is http://app.hnkjt.gov.cn/web/index.do(accessed on 09 February 2025).
文摘The photovoltaic(PV)output process is inherently complex,often disrupted by a multitude of meteoro-logical factors,while conventional detection methods at PV power stations prove inadequate,compromising prediction accuracy.To address this challenge,this paper introduces a power prediction method that leverages modal switching(MS),weight factor adjustment(WFA),and parallel long short-term memory(PALSTM).Initially,historical PV power station data is categorized into distinct modes based on global horizontal irradiance and converted solar angles.Correlation analysis is then employed to evaluate the impact of various meteorological factors on PV power,selecting those with strong correlations for each specific mode.Subsequently,the weights of meteorological parameters are optimized and adjusted,and a PALSTM neural network is constructed,with its parallel modal parameters refined through training.Depending on the prediction time and input data mode characteristics,the appropriate mode channel is selected to forecast PV power station generation.Ultimately,the feasibility of this method is validated through an illustrative analysis of measured data from an Australian PV power station.Comparative test results underscore the method’s advantages,particularly in scenarios where existing detection methods are lacking and meteorological factors frequently fluctuate,demonstrating its superior prediction accuracy and stability.
基金Project(51175287)supported by the National Natural Science Foundation of ChinaProject(2006AA04Z112)supported by National High Technology Research and Development Program of China
文摘A hierarchical scheme of feature-based model similarity measurement was proposed,named CSG_D2,in which both geometry similarity and topology similarity were applied.The features of 3D mechanical part were constructed by a series of primitive features with tree structure,as a form of constructive solid geometry(CSG) tree.The D2 shape distributions of these features were extracted for geometry similarity measurement,and the pose vector and non-disappeared proportion of each leaf node were gained for topology similarity measurement.Based on these,the dissimilarity between the query and the candidate was accessed by level-by-level CSG tree comparisons.With the adjustable weights,our scheme satisfies different comparison emphasis on the geometry or topology similarity.The assessment results from CSG_D2 demonstrate more discriminative than those from D2 in the analysis of precision-recall and similarity matrix.Finally,an experimental search engine is applied for mechanical parts reuse by using CSG_D2,which is convenient for the mechanical design process.
基金Supported by the Program for Yangtse River Scholars and Innovative Research Terms in Universities(IRT0936)the National Basic Research Program of China(2009CB219905+2 种基金2009CB219907)the Daqing Oilfield Co.Ltd
文摘The separate-layer injection in different interlayers and the injection of the same-molecular-weight polymer so- lution in a layer are necessary in the polymer flooding process because of heterogeneous multilayer sandstone reservoirs in EOR projects. To alleviate the matching problems between the layer permeability and the injected polymer molecular weight, a molecular weight adjusting device with porous medium was designed on the basis of mechanical degradation principle. In terms of four variables （polymer concentration, pore diameter, length of shear component and flow rate ）, the theological behavior of hydrolyzed polyacrylamide （HPAM） solu- tion flowing through the device was investigated in detail. The change of these variables is able to control the shear rate of HPAM solutions through ceramic foam, and achieve the desired degree of shear degradation and the final theological parameters-viscosity loss, viscoelasticity and pressure drop. Therefore, a linear relationship between viscosity loss and shearing rate was established so as to obtain the targeted viscosity easily. Field tests in the Daqing Oil Field showed that the polymer molecular weight could drop 20% to 50%. In a word, the results could guide the industrial application of the novel device and the further study of polymer degradation flowing through the porous medium.
基金supported by the National Natural Science Foundation of China(Nos.62002100,61902237)Key Research and Promotion Projects of Henan Province(Nos.232102240023,232102210063,222102210040).
文摘Fingerprint features,as unique and stable biometric identifiers,are crucial for identity verification.However,traditional centralized methods of processing these sensitive data linked to personal identity pose significant privacy risks,potentially leading to user data leakage.Federated Learning allows multiple clients to collaboratively train and optimize models without sharing raw data,effectively addressing privacy and security concerns.However,variations in fingerprint data due to factors such as region,ethnicity,sensor quality,and environmental conditions result in significant heterogeneity across clients.This heterogeneity adversely impacts the generalization ability of the global model,limiting its performance across diverse distributions.To address these challenges,we propose an Adaptive Federated Fingerprint Recognition algorithm(AFFR)based on Federated Learning.The algorithm incorporates a generalization adjustment mechanism that evaluates the generalization gap between the local models and the global model,adaptively adjusting aggregation weights to mitigate the impact of heterogeneity caused by differences in data quality and feature characteristics.Additionally,a noise mechanism is embedded in client-side training to reduce the risk of fingerprint data leakage arising from weight disclosures during model updates.Experiments conducted on three public datasets demonstrate that AFFR significantly enhances model accuracy while ensuring robust privacy protection,showcasing its strong application potential and competitiveness in heterogeneous data environments.
基金supported by the Chinese Ministry of Science and Intergovernmental Cooperation Project (2009DFA12870)the National Science Foundation of China (60974062,60972119)
文摘This paper proposes a particle swarm optimization（PSO） based particle filter（PF） tracking framework,the embedded PSO makes particles move toward the high likelihood area to find the optimal position in the state transition stage,and simultaneously incorporates the newest observations into the proposal distribution in the update stage.In the proposed approach,likelihood measure functions involving multiple features are presented to enhance the performance of model fitting.Furthermore,the multi-feature weights are self-adaptively adjusted by a PSO algorithm throughout the tracking process.There are three main contributions.Firstly,the PSO algorithm is fused into the PF framework,which can efficiently alleviate the particles degeneracy phenomenon.Secondly,an effective convergence criterion for the PSO algorithm is explored,which can avoid particles getting stuck in local minima and maintain a greater particle diversity.Finally,a multi-feature weight self-adjusting strategy is proposed,which can significantly improve the tracking robustness and accuracy.Experiments performed on several challenging public video sequences demonstrate that the proposed tracking approach achieves a considerable performance.
文摘Nowadays, internet-based surveys are increasingly used for data collection, because their usage is simple and cheap. Also they give fast access to a large group of respondents. There are many factors affecting internet surveys, such as measurement, survey design and sampling selection bias. The sampling has an important place in selection bias in internet survey. In terms of sample selection, the type of access to internet surveys has several limitations. There are internet surveys based on restricted access and on voluntary participation, and these are characterized by their implementation according to the type of survey. It can be used probability and non-probability sampling, both of which may lead to biased estimates. There are different ways to correct for selection biases;poststratification or weighting class adjustments, raking or rim weighting, generalized regression modeling and propensity score adjustments. This paper aims to describe methodological problems about selection bias issues and to give a review in internet surveys. Also the objective of this study is to show the effect of various correction techniques for reducing selection bias.
基金supported by the Natural Science Foundation of China under Grant Nos.10771017 and 11071022Key Project of MOE,PRC under Grant No.309007
文摘The multivariate linear errors-in-variables model when the regressors are missing at random in the sense of Rubin （1976） is considered in this paper. A constrained empirical likelihood confidence region for a parameter β0 in this model is proposed, which is constructed by combining the score function corresponding to the weighted squared orthogonal distance based on inverse probability with a constrained region of β0. It is shown that the empirical log-likelihood ratio at the true parameter converges to the standard chi-square distribution. Simulations show that the coverage rate of the proposed confidence region is closer to the nominal level and the length of confidence interval is narrower than those of the normal approximation of inverse probability weighted adjusted least square estimator in most cases. A real example is studied and the result supports the theory and simulation＇s conclusion.
基金Supported by the Educational Commission of Hubei Province of China(Grant No.D20112503)National Natural Science Foundation of China(Grant Nos.11071022,11231010 and 11028103)the foundation of Beijing Center of Mathematics and Information Sciences
文摘In this paper, the parameters of a p-dimensional linear structural EV(error-in-variable)model are estimated when the coefficients vary with a real variable and the model error is time series.The adjust weighted least squares(AWLS) method is used to estimate the parameters. It is shown that the estimators are weakly consistent and asymptotically normal, and the optimal convergence rate is also obtained. Simulations study are undertaken to illustrate our AWLSEs have good performance.
