While progress has been made in information source localization,it has overlooked the prevalent friend and adversarial relationships in social networks.This paper addresses this gap by focusing on source localization ...While progress has been made in information source localization,it has overlooked the prevalent friend and adversarial relationships in social networks.This paper addresses this gap by focusing on source localization in signed network models.Leveraging the topological characteristics of signed networks and transforming the propagation probability into effective distance,we propose an optimization method for observer selection.Additionally,by using the reverse propagation algorithm we present a method for information source localization in signed networks.Extensive experimental results demonstrate that a higher proportion of positive edges within signed networks contributes to more favorable source localization,and the higher the ratio of propagation rates between positive and negative edges,the more accurate the source localization becomes.Interestingly,this aligns with our observation that,in reality,the number of friends tends to be greater than the number of adversaries,and the likelihood of information propagation among friends is often higher than among adversaries.In addition,the source located at the periphery of the network is not easy to identify.Furthermore,our proposed observer selection method based on effective distance achieves higher operational efficiency and exhibits higher accuracy in information source localization,compared with three strategies for observer selection based on the classical full-order neighbor coverage.展开更多
To remedy the empirical pitfalls of current chinese specifications and MUTCD 2009 guidelines in determining the placement distance of freeway exit advance guide signs,the driving maneuver of exiting traffic is analyze...To remedy the empirical pitfalls of current chinese specifications and MUTCD 2009 guidelines in determining the placement distance of freeway exit advance guide signs,the driving maneuver of exiting traffic is analyzed and the factors influencing placement distance are explored.Variables including the number of lanes,lane width,lane-changing time,driver＇s visual characteristics,sign installation methods and operating speeds on both freeway mainlines and exit ramps are found significant in explaining exit safety.Three different installation methods,namely ground installation,overhead installation and median installation,are introduced and their applicable conditions are given.Models,with the same structure among the three installation methods,are developed to compute the placement distance under different roadway geometric and traffic conditions.Taking overhead installation as an example,simulation results in TSIS-CORSIM show that the proposed distance reduces the number of lane changes in the area from the ramp nose to 500 m upstream by 58.93% compared with current Chinese specifications and 27.35% compared with MUTCD 2009 guidelines.Thus,the distances recommended in this paper have a better safety performance.展开更多
Sign language recognition is vital for enhancing communication accessibility among the Deaf and hard-of-hearing communities.In Japan,approximately 360,000 individualswith hearing and speech disabilities rely on Japane...Sign language recognition is vital for enhancing communication accessibility among the Deaf and hard-of-hearing communities.In Japan,approximately 360,000 individualswith hearing and speech disabilities rely on Japanese Sign Language(JSL)for communication.However,existing JSL recognition systems have faced significant performance limitations due to inherent complexities.In response to these challenges,we present a novel JSL recognition system that employs a strategic fusion approach,combining joint skeleton-based handcrafted features and pixel-based deep learning features.Our system incorporates two distinct streams:the first stream extracts crucial handcrafted features,emphasizing the capture of hand and body movements within JSL gestures.Simultaneously,a deep learning-based transfer learning stream captures hierarchical representations of JSL gestures in the second stream.Then,we concatenated the critical information of the first stream and the hierarchy of the second stream features to produce the multiple levels of the fusion features,aiming to create a comprehensive representation of the JSL gestures.After reducing the dimensionality of the feature,a feature selection approach and a kernel-based support vector machine(SVM)were used for the classification.To assess the effectiveness of our approach,we conducted extensive experiments on our Lab JSL dataset and a publicly available Arabic sign language(ArSL)dataset.Our results unequivocally demonstrate that our fusion approach significantly enhances JSL recognition accuracy and robustness compared to individual feature sets or traditional recognition methods.展开更多
Human activity detection and recognition is a challenging task.Video surveillance can benefit greatly by advances in Internet of Things(IoT)and cloud computing.Artificial intelligence IoT(AIoT)based devices form the b...Human activity detection and recognition is a challenging task.Video surveillance can benefit greatly by advances in Internet of Things(IoT)and cloud computing.Artificial intelligence IoT(AIoT)based devices form the basis of a smart city.The research presents Intelligent dynamic gesture recognition(IDGR)using a Convolutional neural network(CNN)empowered by edit distance for video recognition.The proposed system has been evaluated using AIoT enabled devices for static and dynamic gestures of Pakistani sign language(PSL).However,the proposed methodology can work efficiently for any type of video.The proposed research concludes that deep learning and convolutional neural networks give a most appropriate solution retaining discriminative and dynamic information of the input action.The research proposes recognition of dynamic gestures using image recognition of the keyframes based on CNN extracted from the human activity.Edit distance is used to find out the label of the word to which those sets of frames belong to.The simulation results have shown that at 400 videos per human action,100 epochs,234×234 image size,the accuracy of the system is 90.79%,which is a reasonable accuracy for a relatively small dataset as compared to the previously published techniques.展开更多
This paper combines interval-valued intuitionistic fuzzy sets and rough sets.It studies rougheness in interval-valued intuitionistic fuzzy sets and proposes one kind of interval-valued intuitionistic fuzzy-rough sets ...This paper combines interval-valued intuitionistic fuzzy sets and rough sets.It studies rougheness in interval-valued intuitionistic fuzzy sets and proposes one kind of interval-valued intuitionistic fuzzy-rough sets models under the equivalence relation in crisp sets.That extends the classical rough set defined by Pawlak.展开更多
This paper studies the geometric boundary representations for Inverse Lax-Wendroff(ILW)method,aiming to develop a practical computer-aided engineering method without body-fitted meshes.We propose the signed distance f...This paper studies the geometric boundary representations for Inverse Lax-Wendroff(ILW)method,aiming to develop a practical computer-aided engineering method without body-fitted meshes.We propose the signed distance function(SDF)representation of the geometric boundary and design an extremely efficient algorithm for foot point calculation,which is particularly in line with the needs of ILW.Theoretical and numerical analyses demonstrate that the SDF representation of geometric boundary can satisfy ILW’s needs better than others.The effectiveness and robustness of our proposed method are verified by simulating initial boundary value computational physical problems of Euler equation for compressible fluids.展开更多
Traffic intersections are incredibly dangerous for drivers and pedestrians. Statistics from both Canada and the U.S. show a high number of fatalities and serious injuries related to crashes at intersections. In Canada...Traffic intersections are incredibly dangerous for drivers and pedestrians. Statistics from both Canada and the U.S. show a high number of fatalities and serious injuries related to crashes at intersections. In Canada, during 2019, the National Collision Database shows that 28% of traffic fatalities and 42% of serious injuries occurred at intersections. Likewise, the U.S. National Highway Traffic Administration (NHTSA) found that about 40% of the estimated 5,811,000 accidents in the U.S. during the year studied were intersection-related crashes. In fact, a major survey by the car insurance industry found that nearly 85% of drivers could not identify the correct action to take when approaching a yellow traffic light at an intersection. One major reason for these accidents is the “yellow light dilemma,” the ambiguous situation where a driver should stop or proceed forward when unexpectedly faced with a yellow light. This situation is even further exacerbated by the tendency of aggressive drivers to inappropriately speed up on the yellow just to get through the traffic light. A survey of Canadian drivers conducted by the Traffic Injury Research Foundation found that 9% of drivers admitted to speeding up to get through a traffic light. Another reason for these accidents is the increased danger of making a left-hand turn on yellow. According to the National Highway Traffic Safety Association (NHTSA), left turns occur in approximately 22.2% of collisions—as opposed to just 1.2% for right turns. Moreover, a study by CNN found left turns are three times as likely to kill pedestrians than right turns. The reason left turns are so much more likely to cause an accident is because they take a driver against traffic and in the path of oncoming cars. Additionally, most of these left turns occur at the driver’s discretion—as opposed to the distressingly brief left-hand arrow at busy intersections. Drive Safe Now proposes a workable solution for reducing the number of accidents occurring during a yellow light at intersections. We believe this fairly simple solution will save lives, prevent injuries, reduce damage to public and private property, and decrease insurance costs.展开更多
将神经网络用于场景几何材质的高效表达,结合逆向渲染在二维光度图的监督下重建高质量的网格和材质贴图,为现有的图形学流水线提供服务——神经渲染已成为近年来计算机图形学新的研究热点。在IRON(inverse rendering by optimizing neur...将神经网络用于场景几何材质的高效表达,结合逆向渲染在二维光度图的监督下重建高质量的网格和材质贴图,为现有的图形学流水线提供服务——神经渲染已成为近年来计算机图形学新的研究热点。在IRON(inverse rendering by optimizing neural SDFs and materials from photometric images)神经渲染模型基础上,通过引入多分辨率哈希编码,采用冻结训练等方法提高原始模型的训练速度。在多个数据集上的对比实验表明,优化后的IRON逆渲染模型训练速度提升了约40%,且重建结果中包含更多细节。展开更多
基金Project supported by the National Natural Science Foundation of China(Grant Nos.62103375 and 62006106)the Zhejiang Provincial Philosophy and Social Science Planning Project(Grant No.22NDJC009Z)+1 种基金the Education Ministry Humanities and Social Science Foundation of China(Grant Nos.19YJCZH056 and 21YJC630120)the Natural Science Foundation of Zhejiang Province of China(Grant Nos.LY23F030003 and LQ21F020005).
文摘While progress has been made in information source localization,it has overlooked the prevalent friend and adversarial relationships in social networks.This paper addresses this gap by focusing on source localization in signed network models.Leveraging the topological characteristics of signed networks and transforming the propagation probability into effective distance,we propose an optimization method for observer selection.Additionally,by using the reverse propagation algorithm we present a method for information source localization in signed networks.Extensive experimental results demonstrate that a higher proportion of positive edges within signed networks contributes to more favorable source localization,and the higher the ratio of propagation rates between positive and negative edges,the more accurate the source localization becomes.Interestingly,this aligns with our observation that,in reality,the number of friends tends to be greater than the number of adversaries,and the likelihood of information propagation among friends is often higher than among adversaries.In addition,the source located at the periphery of the network is not easy to identify.Furthermore,our proposed observer selection method based on effective distance achieves higher operational efficiency and exhibits higher accuracy in information source localization,compared with three strategies for observer selection based on the classical full-order neighbor coverage.
基金Project of Florida Department of Transportation(No.BD54438)the National Key Technology R&D Program of China during the 11th Five-Year Plan Period(No.2006BAJ18B03)
文摘To remedy the empirical pitfalls of current chinese specifications and MUTCD 2009 guidelines in determining the placement distance of freeway exit advance guide signs,the driving maneuver of exiting traffic is analyzed and the factors influencing placement distance are explored.Variables including the number of lanes,lane width,lane-changing time,driver＇s visual characteristics,sign installation methods and operating speeds on both freeway mainlines and exit ramps are found significant in explaining exit safety.Three different installation methods,namely ground installation,overhead installation and median installation,are introduced and their applicable conditions are given.Models,with the same structure among the three installation methods,are developed to compute the placement distance under different roadway geometric and traffic conditions.Taking overhead installation as an example,simulation results in TSIS-CORSIM show that the proposed distance reduces the number of lane changes in the area from the ramp nose to 500 m upstream by 58.93% compared with current Chinese specifications and 27.35% compared with MUTCD 2009 guidelines.Thus,the distances recommended in this paper have a better safety performance.
基金supported by the Competitive Research Fund of the University of Aizu,Japan.
文摘Sign language recognition is vital for enhancing communication accessibility among the Deaf and hard-of-hearing communities.In Japan,approximately 360,000 individualswith hearing and speech disabilities rely on Japanese Sign Language(JSL)for communication.However,existing JSL recognition systems have faced significant performance limitations due to inherent complexities.In response to these challenges,we present a novel JSL recognition system that employs a strategic fusion approach,combining joint skeleton-based handcrafted features and pixel-based deep learning features.Our system incorporates two distinct streams:the first stream extracts crucial handcrafted features,emphasizing the capture of hand and body movements within JSL gestures.Simultaneously,a deep learning-based transfer learning stream captures hierarchical representations of JSL gestures in the second stream.Then,we concatenated the critical information of the first stream and the hierarchy of the second stream features to produce the multiple levels of the fusion features,aiming to create a comprehensive representation of the JSL gestures.After reducing the dimensionality of the feature,a feature selection approach and a kernel-based support vector machine(SVM)were used for the classification.To assess the effectiveness of our approach,we conducted extensive experiments on our Lab JSL dataset and a publicly available Arabic sign language(ArSL)dataset.Our results unequivocally demonstrate that our fusion approach significantly enhances JSL recognition accuracy and robustness compared to individual feature sets or traditional recognition methods.
文摘Human activity detection and recognition is a challenging task.Video surveillance can benefit greatly by advances in Internet of Things(IoT)and cloud computing.Artificial intelligence IoT(AIoT)based devices form the basis of a smart city.The research presents Intelligent dynamic gesture recognition(IDGR)using a Convolutional neural network(CNN)empowered by edit distance for video recognition.The proposed system has been evaluated using AIoT enabled devices for static and dynamic gestures of Pakistani sign language(PSL).However,the proposed methodology can work efficiently for any type of video.The proposed research concludes that deep learning and convolutional neural networks give a most appropriate solution retaining discriminative and dynamic information of the input action.The research proposes recognition of dynamic gestures using image recognition of the keyframes based on CNN extracted from the human activity.Edit distance is used to find out the label of the word to which those sets of frames belong to.The simulation results have shown that at 400 videos per human action,100 epochs,234×234 image size,the accuracy of the system is 90.79%,which is a reasonable accuracy for a relatively small dataset as compared to the previously published techniques.
基金supported by grants from the National Natural Science Foundation of China(Nos.10971185 and 10971186)the Natural Science Foundation of Fujiang Province in China(No.2008F5066).
文摘This paper combines interval-valued intuitionistic fuzzy sets and rough sets.It studies rougheness in interval-valued intuitionistic fuzzy sets and proposes one kind of interval-valued intuitionistic fuzzy-rough sets models under the equivalence relation in crisp sets.That extends the classical rough set defined by Pawlak.
文摘This paper studies the geometric boundary representations for Inverse Lax-Wendroff(ILW)method,aiming to develop a practical computer-aided engineering method without body-fitted meshes.We propose the signed distance function(SDF)representation of the geometric boundary and design an extremely efficient algorithm for foot point calculation,which is particularly in line with the needs of ILW.Theoretical and numerical analyses demonstrate that the SDF representation of geometric boundary can satisfy ILW’s needs better than others.The effectiveness and robustness of our proposed method are verified by simulating initial boundary value computational physical problems of Euler equation for compressible fluids.
文摘Traffic intersections are incredibly dangerous for drivers and pedestrians. Statistics from both Canada and the U.S. show a high number of fatalities and serious injuries related to crashes at intersections. In Canada, during 2019, the National Collision Database shows that 28% of traffic fatalities and 42% of serious injuries occurred at intersections. Likewise, the U.S. National Highway Traffic Administration (NHTSA) found that about 40% of the estimated 5,811,000 accidents in the U.S. during the year studied were intersection-related crashes. In fact, a major survey by the car insurance industry found that nearly 85% of drivers could not identify the correct action to take when approaching a yellow traffic light at an intersection. One major reason for these accidents is the “yellow light dilemma,” the ambiguous situation where a driver should stop or proceed forward when unexpectedly faced with a yellow light. This situation is even further exacerbated by the tendency of aggressive drivers to inappropriately speed up on the yellow just to get through the traffic light. A survey of Canadian drivers conducted by the Traffic Injury Research Foundation found that 9% of drivers admitted to speeding up to get through a traffic light. Another reason for these accidents is the increased danger of making a left-hand turn on yellow. According to the National Highway Traffic Safety Association (NHTSA), left turns occur in approximately 22.2% of collisions—as opposed to just 1.2% for right turns. Moreover, a study by CNN found left turns are three times as likely to kill pedestrians than right turns. The reason left turns are so much more likely to cause an accident is because they take a driver against traffic and in the path of oncoming cars. Additionally, most of these left turns occur at the driver’s discretion—as opposed to the distressingly brief left-hand arrow at busy intersections. Drive Safe Now proposes a workable solution for reducing the number of accidents occurring during a yellow light at intersections. We believe this fairly simple solution will save lives, prevent injuries, reduce damage to public and private property, and decrease insurance costs.
文摘将神经网络用于场景几何材质的高效表达,结合逆向渲染在二维光度图的监督下重建高质量的网格和材质贴图,为现有的图形学流水线提供服务——神经渲染已成为近年来计算机图形学新的研究热点。在IRON(inverse rendering by optimizing neural SDFs and materials from photometric images)神经渲染模型基础上,通过引入多分辨率哈希编码,采用冻结训练等方法提高原始模型的训练速度。在多个数据集上的对比实验表明,优化后的IRON逆渲染模型训练速度提升了约40%,且重建结果中包含更多细节。
