Shape prediction of deformable linear objects(DLO)plays critical roles in robotics,medical devices,aerospace,and manufacturing,especially in manipulating objects such as cables,wires,and fibers.Due to the inherent fle...Shape prediction of deformable linear objects(DLO)plays critical roles in robotics,medical devices,aerospace,and manufacturing,especially in manipulating objects such as cables,wires,and fibers.Due to the inherent flexibility of DLO and their complex deformation behaviors,such as bending and torsion,it is challenging to predict their dynamic characteristics accurately.Although the traditional physical modeling method can simulate the complex deformation behavior of DLO,the calculation cost is high and it is difficult to meet the demand of real-time prediction.In addition,the scarcity of data resources also limits the prediction accuracy of existing models.To solve these problems,a method of fiber shape prediction based on a physical information graph neural network(PIGNN)is proposed in this paper.This method cleverly combines the powerful expressive power of graph neural networks with the strict constraints of physical laws.Specifically,we learn the initial deformation model of the fiber through graph neural networks(GNN)to provide a good initial estimate for the model,which helps alleviate the problem of data resource scarcity.During the training process,we incorporate the physical prior knowledge of the dynamic deformation of the fiber optics into the loss function as a constraint,which is then fed back to the network model.This ensures that the shape of the fiber optics gradually approaches the true target shape,effectively solving the complex nonlinear behavior prediction problem of deformable linear objects.Experimental results demonstrate that,compared to traditional methods,the proposed method significantly reduces execution time and prediction error when handling the complex deformations of deformable fibers.This showcases its potential application value and superiority in fiber manipulation.展开更多
A literature analysis has shown that object search,recognition,and tracking systems are becoming increasingly popular.However,such systems do not achieve high practical results in analyzing small moving living objects...A literature analysis has shown that object search,recognition,and tracking systems are becoming increasingly popular.However,such systems do not achieve high practical results in analyzing small moving living objects ranging from 8 to 14 mm.This article examines methods and tools for recognizing and tracking the class of small moving objects,such as ants.To fulfill those aims,a customized You Only Look Once Ants Recognition(YOLO_AR)Convolutional Neural Network(CNN)has been trained to recognize Messor Structor ants in the laboratory using the LabelImg object marker tool.The proposed model is an extension of the You Only Look Once v4(Yolov4)512×512 model with an additional Self Regularized Non–Monotonic(Mish)activation function.Additionally,the scalable solution for continuous object recognizing and tracking was implemented.This solution is based on the OpenDatacam system,with extended Object Tracking modules that allow for tracking and counting objects that have crossed the custom boundary line.During the study,the methods of the alignment algorithm for finding the trajectory of moving objects were modified.I discovered that the Hungarian algorithm showed better results in tracking small objects than the K–D dimensional tree(k-d tree)matching algorithm used in OpenDataCam.Remarkably,such an algorithm showed better results with the implemented YOLO_AR model due to the lack of False Positives(FP).Therefore,I provided a new tracker module with a Hungarian matching algorithm verified on the Multiple Object Tracking(MOT)benchmark.Furthermore,additional customization parameters for object recognition and tracking results parsing and filtering were added,like boundary angle threshold(BAT)and past frames trajectory prediction(PFTP).Experimental tests confirmed the results of the study on a mobile device.During the experiment,parameters such as the quality of recognition and tracking of moving objects,the PFTP and BAT,and the configuration parameters of the neural network and boundary line model were analyzed.The results showed an increased tracking accuracy with the proposed methods by 50%.The study results confirmed the relevance of the topic and the effectiveness of the implemented methods and tools.展开更多
Dear Editor,This letter proposes a novel Nash bargaining solution-based multiobjective model predictive control(MPC)scheme to deal with the interaction force control and the path-following problem of the constrained i...Dear Editor,This letter proposes a novel Nash bargaining solution-based multiobjective model predictive control(MPC)scheme to deal with the interaction force control and the path-following problem of the constrained interactive robot.Considering the elastic interaction force model,a mechanical trade-off always exists between the interaction force and position,which means that neither force nor path following can satisfy their desired demands completely.Based on this consideration,two irreconcilable control specifications,the force object function and the position track object function,are proposed,and a new multi-objective MPC scheme is then designed.展开更多
In the new era of technology,daily human activities are becoming more challenging in terms of monitoring complex scenes and backgrounds.To understand the scenes and activities from human life logs,human-object interac...In the new era of technology,daily human activities are becoming more challenging in terms of monitoring complex scenes and backgrounds.To understand the scenes and activities from human life logs,human-object interaction(HOI)is important in terms of visual relationship detection and human pose estimation.Activities understanding and interaction recognition between human and object along with the pose estimation and interaction modeling have been explained.Some existing algorithms and feature extraction procedures are complicated including accurate detection of rare human postures,occluded regions,and unsatisfactory detection of objects,especially small-sized objects.The existing HOI detection techniques are instancecentric(object-based)where interaction is predicted between all the pairs.Such estimation depends on appearance features and spatial information.Therefore,we propose a novel approach to demonstrate that the appearance features alone are not sufficient to predict the HOI.Furthermore,we detect the human body parts by using the Gaussian Matric Model(GMM)followed by object detection using YOLO.We predict the interaction points which directly classify the interaction and pair them with densely predicted HOI vectors by using the interaction algorithm.The interactions are linked with the human and object to predict the actions.The experiments have been performed on two benchmark HOI datasets demonstrating the proposed approach.展开更多
There is a huge demand to develop a method for marine search and rescue（SAR） operators automatically predicting the most probable searching area of the drifting object. This paper presents a novel drifting predictio...There is a huge demand to develop a method for marine search and rescue（SAR） operators automatically predicting the most probable searching area of the drifting object. This paper presents a novel drifting prediction model to improve the accuracy of the drifting trajectory computation of the sea-surface objects. First, a new drifting kinetic model based on the geometry characteristics of the objects is proposed that involves the effects of the object shape and stochastic motion features in addition to the traditional factors of wind and currents. Then, a computer simulation-based method is employed to analyze the stochastic motion features of the drifting objects, which is applied to estimate the uncertainty parameters of the stochastic factors of the drifting objects. Finally, the accuracy of the model is evaluated by comparison with the flume experimental results. It is shown that the proposed method can be used for various shape objects in the drifting trajectory prediction and the maritime search and rescue decision-making system.展开更多
Recent advances in supervised salient object detection modeling has resulted in significant performance improvements on benchmark datasets. However, most of the existing salient object detection models assume that at ...Recent advances in supervised salient object detection modeling has resulted in significant performance improvements on benchmark datasets. However, most of the existing salient object detection models assume that at least one salient object exists in the input image. Such an assumption often leads to less appealing saliency maps on the background images with no salient object at all. Therefore, handling those cases can reduce the false positive rate of a model. In this paper, we propose a supervised learning approach for jointly addressing the salient object detection and existence prediction problems. Given a set of background-only images and images with salient objects, as well as their salient object annotations, we adopt the structural SVM framework and formulate the two problems jointly in a single integrated objective function: saliency labels of superpixels are involved in a classification term conditioned on the salient object existence variable, which in turn depends on both global image and regional saliency features and saliency labels assignments. The loss function also considers both image-level and regionlevel mis-classifications. Extensive evaluation on benchmark datasets validate the effectiveness of our proposed joint approach compared to the baseline and state-of-the-art models.展开更多
This paper discusses smart body sensor objects （BSOs）, including their networking and internetworking. Smartness can be incorpo-rated into BSOs by embedding virtualization, predictive analytics, and proactive comput...This paper discusses smart body sensor objects （BSOs）, including their networking and internetworking. Smartness can be incorpo-rated into BSOs by embedding virtualization, predictive analytics, and proactive computing and communications capabilities. A few use cases including the relevant privacy and protocol requirements are also presented. General usage and deployment eti-quette along with the relevant regulatory implications are then discussed.展开更多
Emerging technologies of wireless and mobile communication enable people to accumulate a large volume of time-stamped locations,which appear in the form of a continuous moving object trajectory.How to accurately predi...Emerging technologies of wireless and mobile communication enable people to accumulate a large volume of time-stamped locations,which appear in the form of a continuous moving object trajectory.How to accurately predict the uncertain mobility of objects becomes an important and challenging problem.Existing algorithms for trajectory prediction in moving objects databases mainly focus on identifying frequent trajectory patterns,and do not take account of the effect of essential dynamic environmental factors.In this study,a general schema for predicting uncertain trajectories of moving objects with dynamic environment awareness is presented,and the key techniques in trajectory prediction arc addressed in detail.In order to accurately predict the trajectories,a trajectory prediction algorithm based on continuous time Bayesian networks（CTBNs） is improved and applied,which takes dynamic environmental factors into full consideration.Experiments conducted on synthetic trajectory data verify the effectiveness of the improved algorithm,which also guarantees the time performance as well.展开更多
Interval prediction of wind power,which features the upper and lower limits of wind power at a given confidence level,plays a significant role in accurate prediction and stability of the power grid integrated with win...Interval prediction of wind power,which features the upper and lower limits of wind power at a given confidence level,plays a significant role in accurate prediction and stability of the power grid integrated with wind power.However,the conventional methods of interval prediction are commonly based on a hypothetic probability distribution function,which neglects the correlations among various variables,leading to the decrease of prediction accuracy.Therefore,we improve the multi-objective interval prediction based on the conditional copula function,through which we can fully utilize the correlations among variables to improve prediction accuracy without an assumed probability distribution function.We use the multi-objective optimization method of nondominated sorting genetic algorithm-II(NSGA-II)to obtain the optimal solution set.The particular best solution is weighted by the prediction interval average width(PIAW)and prediction interval coverage probability(PICP)to pick the optimized solution in practical examples.Finally,we apply the proposed method to three wind power plants in different cities in China as examples forvalidation and obtain higher prediction accuracy compared with other methods,i.e.,relevance vector machine(RVM),artificial neural network(ANN),and particle swarm optimization kernel extreme learning machine(PSO-KELM).These results demonstrate the superiority and practicability of this method in interval prediction of wind power.展开更多
When deriving the Fourier diffraction theorem based on the first-order Born approximation,the difference between wave number of the scattering object and that of the surrounding medium is ignored,causing substantial e...When deriving the Fourier diffraction theorem based on the first-order Born approximation,the difference between wave number of the scattering object and that of the surrounding medium is ignored,causing substantial errors in sound scattering prediction.This paper modifies the Born approximation by taking into account the amplitude and phase changes between the scattering object and the water due to the wave number difference.By changing the radius and center position of the sampling circle in the Fourier domain,accuracy of the predicted sound scattering is improved.With the modified Born approximation,the computed far-field directional pattern of the scattered sound from a circular cylinder is in good agreement with the rigorous solution.Numerical calculations for several objects with different shapes are used to show applicability and effectiveness of the proposed method.展开更多
Augmented virtual environments(AVE)combine real-time videos with 3D scenes in a Digital Earth System or 3D GIS to present dynamic information and a virtual scene simultaneously.AVE can provide solutions for continuous...Augmented virtual environments(AVE)combine real-time videos with 3D scenes in a Digital Earth System or 3D GIS to present dynamic information and a virtual scene simultaneously.AVE can provide solutions for continuous tracking of moving objects,camera scheduling,and path planning in the real world.This paper proposes a novel approach for 3D path prediction of moving objects in a video-augmented indoor virtual environment.The study includes 3D motion analysis of moving objects,multi-path prediction,hierarchical visualization,and path-based multi-camera scheduling.The results show that these methods can give a closed-loop process of 3D path prediction and continuous tracking of moving objects in an AVE.The path analysis algorithms proved accurate and time-efficient,costing less than 1.3 ms to get the optimal path.The experiment ran a 3D scene containing 295,000 triangles at around 35 frames per second on a laptop with 1 GB of graphics card memory,which means the performance of the proposed methods is good enough to maintain high rendering efficiency for a video-augmented indoor virtual scene.展开更多
针对通用目标检测算法在检测小目标时存在错检和漏检等问题,提出了一种小目标检测算法IPH(Involution Prediction Head),将其运用在YOLOv4和YOLOv5的检测头部分,在VOC2007数据集上的实验结果表明,运用IPH后的YOLOv4小目标检测精度APs(AP...针对通用目标检测算法在检测小目标时存在错检和漏检等问题,提出了一种小目标检测算法IPH(Involution Prediction Head),将其运用在YOLOv4和YOLOv5的检测头部分,在VOC2007数据集上的实验结果表明,运用IPH后的YOLOv4小目标检测精度APs(AP for small objects)相比原始算法提升了1.1%,在YOLOv5上的APs更是提升了5.9%。经智能交通检测数据集进一步检验,IPH算法和去下采样能有效提升小目标检测精度,减少错检和漏检的情况。展开更多
Background Co-salient object detection(Co-SOD)aims to identify and segment commonly salient objects in a set of related images.However,most current Co-SOD methods encounter issues with the inclusion of irrelevant info...Background Co-salient object detection(Co-SOD)aims to identify and segment commonly salient objects in a set of related images.However,most current Co-SOD methods encounter issues with the inclusion of irrelevant information in the co-representation.These issues hamper their ability to locate co-salient objects and significantly restrict the accuracy of detection.Methods To address this issue,this study introduces a novel Co-SOD method with iterative purification and predictive optimization(IPPO)comprising a common salient purification module(CSPM),predictive optimizing module(POM),and diminishing mixed enhancement block(DMEB).Results These components are designed to explore noise-free joint representations,assist the model in enhancing the quality of the final prediction results,and significantly improve the performance of the Co-SOD algorithm.Furthermore,through a comprehensive evaluation of IPPO and state-of-the-art algorithms focusing on the roles of CSPM,POM,and DMEB,our experiments confirmed that these components are pivotal in enhancing the performance of the model,substantiating the significant advancements of our method over existing benchmarks.Experiments on several challenging benchmark co-saliency datasets demonstrate that the proposed IPPO achieves state-of-the-art performance.展开更多
The determination and optimization of Automatic Train Operation(ATO) control strategy is one of the most critical technologies for urban rail train operation. The practical ATO optimal control strategy must consider m...The determination and optimization of Automatic Train Operation(ATO) control strategy is one of the most critical technologies for urban rail train operation. The practical ATO optimal control strategy must consider many goals of the train operation, such as safety, accuracy, comfort, energy saving and so on. This paper designs a set of efficient and universal multi-objective control strategy. Firstly, based on the analysis of urban rail transit and its operating environment, the multi-objective optimization model considering all the indexes of train operation is established by using multi-objective optimization theory. Secondly, Non-dominated Sorting Genetic Algorithm II(NSGA-II) is used to solve the model, and the optimal speed curve of train running is generated.Finally, the intelligent controller is designed by the combination of fuzzy controller algorithm and the predictive control algorithm, which can control and optimize the train operation in real time. Then the robustness of the control system can ensure and the requirements for multi-objective in train operation can be satisfied.展开更多
High-entropy alloys(HEAs)have emerged as promising candidates for catalyst applications due to their inherent compositional,structural,and site-level diversities,which enable highly tunable catalytic properties.Howeve...High-entropy alloys(HEAs)have emerged as promising candidates for catalyst applications due to their inherent compositional,structural,and site-level diversities,which enable highly tunable catalytic properties.However,these complexities pose grand challenges for traditional“trial-and-error”experimentation or computationally expensive“brute-force”ab initio calculations.Machine learning(ML)demonstrates great potential to address these challenges by establishing efficient,scalable mappings from composition,structure or site environment to HEA properties.Among these properties,adsorption energy,which quantifies the binding strength between catalytic intermediates and surface sites,is a crucial indicator of catalytic activity.This review provides a comprehensive overview ofML-driven strategies for adsorption energy prediction in the context of HEAs.Two primary strategies are introduced:“direct”prediction from unrelaxed structure and“iterative”prediction viaML potential-guided relaxation modeling.Both strategies can leverage handcrafted features or end-toend frameworks such as graph neural networks.We also discuss how pretrained models on largescale databases can extend to out-of-domain HEA systems.Beyond methodology,we address key challenges and future directions,including benchmarking ML strategies,developing HEA-specific datasets,pretraining and fine-tuning,integrating chained ML models,advancing multi-objective optimization,and bridgingMLpredictions with experimental validation.By critically evaluating existing strategies and highlighting emerging trends,this review underscores the critical role of ML in advancing adsorption energy predictions,offering a foundation for accelerating the discovery and optimization of HEA catalysts.展开更多
Energetic Materials(EMs)play important roles in military,civilian and aerospace fields.Energy and stability are the two most important but contradictory properties in practical application,thus leading to difficult ch...Energetic Materials(EMs)play important roles in military,civilian and aerospace fields.Energy and stability are the two most important but contradictory properties in practical application,thus leading to difficult challenges in developing new EMswith high comprehensive performance.Motivated by the challenge,we exploit a de novo design framework targeting multiple objectives by integrating deep learning generator,machine learning prediction models,Pareto front optimization and quantum mechanics(QM)validation.First,heat of explosion(Q)and bond dissociation energy(BDE)are calculated by high-precisionQMfor 778 explosives experimentally reported.With the reliable dataset,RNN coupled with transfer learning is exploited to generate a new massive search space with 2×10^(5)potential energeticmolecules.Qand BDE prediction models with high accuracy are further developed by data augmentation and improvements in feature representation and model architectures,to quickly and accurately evaluate these new energeticmolecules.The modified 3D-GNN achieves an R^(2)=0.95 for the Q prediction,while the XGBoost coupled with the feature complementarity and PADRE data augmentation performs best for the BDE prediction(R2=0.98).To screen energetic compounds with trade-off energy and stability from the vast new molecule space,the predicted values and uncertainties are simultaneously considered,and Pareto front-based multi-objective screening is conducted by using 2D P[I]metric.QM calculation confirms the superior performance of the top 60 candidates to CL-20 in Q.25 promising energetic molecules with high energy and desired stability,as well as synthesis feasibility provide valuable candidates for experimental development.Also,the design strategy can be extended to other material fields.展开更多
Low Young’s modulus and high yield strength are concurrently needed to meet the performance requirements of metallic implant materials.The single-objective performance-oriented alloy design strategies face challenges...Low Young’s modulus and high yield strength are concurrently needed to meet the performance requirements of metallic implant materials.The single-objective performance-oriented alloy design strategies face challenges in effectively addressing the inherent conflict between Young’s modulus and yield strength.In this study,we developed a machine learning model for multi-objective synergistic optimization of modulus and yield strength,successfully enabling simultaneous prediction of Young’s modulus and yield strength in the Ti-Zr-Hf-Nb-Ta-Mo-Sn alloy system.The critical features influencing the modulus and strength of the alloys were systematically analyzed and identified.Moreover,a series of complex concentrated alloy(CCAs)with low Young’s modulus and high yield strength were successfully prepared based on this model.The newly developed alloys exhibited a stable single-phase BCC(body-centered-cubic)structure with Young’s modulus in the range of 40–50 GPa,yield strength of 600–915MPa,and elastic admissible strain of approximately 1.5%.The multi-objective machine learning model developed in this study can synergistically optimize low Young’s modulus and high yield strength in complex alloys,providing a novel approach for the design of advanced biomedical alloys.展开更多
We propose a novel problem revolving around two tasks:(i)given a scene,recommend objects to insert,and(ii)given an object category,retrieve suitable background scenes.A bounding box for the inserted object is predicte...We propose a novel problem revolving around two tasks:(i)given a scene,recommend objects to insert,and(ii)given an object category,retrieve suitable background scenes.A bounding box for the inserted object is predicted in both tasks,which helps downstream applications such as semiautomated advertising and video composition.The major challenge lies in the fact that the target object is neither present nor localized in the input,and furthermore,available datasets only provide scenes with existing objects.To tackle this problem,we build an unsupervised algorithm based on object-level contexts,which explicitly models the joint probability distribution of object categories and bounding boxes using a Gaussian mixture model.Experiments on our own annotated test set demonstrate that our system outperforms existing baselines on all sub-tasks,and does so using a unified framework.Future extensions and applications are suggested.展开更多
In this paper,the recently developed machine learning(ML)approach to improve orbit prediction accuracy is systematically investigated using three ML algorithms,including support vector machine(SVM),artificial neural n...In this paper,the recently developed machine learning(ML)approach to improve orbit prediction accuracy is systematically investigated using three ML algorithms,including support vector machine(SVM),artificial neural network(ANN),and Gaussian processes(GPs).In a simulation environment consisting of orbit propagation,measurement,estimation,and prediction processes,totally 12 resident space objects(RSOs)in solar-synchronous orbit(SSO),low Earth orbit(LEO),and medium Earth orbit(MEO)are simulated to compare the performance of three ML algorithms.The results in this paper show that ANN usually has the best approximation capability but is easiest to overfit data;SVM is the least likely to overfit but the performance usually cannot surpass ANN and GPs.Additionally,the ML approach with all the three algorithms is observed to be robust with respect to the measurement noise.展开更多
基金Supported by the Fundamental Research Funds for the Central Universities(Grant Nos.2232024Y-01,LZB2023001)DHU Distinguished Young Professor Program+1 种基金National Natural Science Foundation of China(Grant No.52275478)AI-Enhanced Research Program of Shanghai Municipal Education Commission(Grant No.SMEC-AI-DHUY-05)。
文摘Shape prediction of deformable linear objects(DLO)plays critical roles in robotics,medical devices,aerospace,and manufacturing,especially in manipulating objects such as cables,wires,and fibers.Due to the inherent flexibility of DLO and their complex deformation behaviors,such as bending and torsion,it is challenging to predict their dynamic characteristics accurately.Although the traditional physical modeling method can simulate the complex deformation behavior of DLO,the calculation cost is high and it is difficult to meet the demand of real-time prediction.In addition,the scarcity of data resources also limits the prediction accuracy of existing models.To solve these problems,a method of fiber shape prediction based on a physical information graph neural network(PIGNN)is proposed in this paper.This method cleverly combines the powerful expressive power of graph neural networks with the strict constraints of physical laws.Specifically,we learn the initial deformation model of the fiber through graph neural networks(GNN)to provide a good initial estimate for the model,which helps alleviate the problem of data resource scarcity.During the training process,we incorporate the physical prior knowledge of the dynamic deformation of the fiber optics into the loss function as a constraint,which is then fed back to the network model.This ensures that the shape of the fiber optics gradually approaches the true target shape,effectively solving the complex nonlinear behavior prediction problem of deformable linear objects.Experimental results demonstrate that,compared to traditional methods,the proposed method significantly reduces execution time and prediction error when handling the complex deformations of deformable fibers.This showcases its potential application value and superiority in fiber manipulation.
文摘A literature analysis has shown that object search,recognition,and tracking systems are becoming increasingly popular.However,such systems do not achieve high practical results in analyzing small moving living objects ranging from 8 to 14 mm.This article examines methods and tools for recognizing and tracking the class of small moving objects,such as ants.To fulfill those aims,a customized You Only Look Once Ants Recognition(YOLO_AR)Convolutional Neural Network(CNN)has been trained to recognize Messor Structor ants in the laboratory using the LabelImg object marker tool.The proposed model is an extension of the You Only Look Once v4(Yolov4)512×512 model with an additional Self Regularized Non–Monotonic(Mish)activation function.Additionally,the scalable solution for continuous object recognizing and tracking was implemented.This solution is based on the OpenDatacam system,with extended Object Tracking modules that allow for tracking and counting objects that have crossed the custom boundary line.During the study,the methods of the alignment algorithm for finding the trajectory of moving objects were modified.I discovered that the Hungarian algorithm showed better results in tracking small objects than the K–D dimensional tree(k-d tree)matching algorithm used in OpenDataCam.Remarkably,such an algorithm showed better results with the implemented YOLO_AR model due to the lack of False Positives(FP).Therefore,I provided a new tracker module with a Hungarian matching algorithm verified on the Multiple Object Tracking(MOT)benchmark.Furthermore,additional customization parameters for object recognition and tracking results parsing and filtering were added,like boundary angle threshold(BAT)and past frames trajectory prediction(PFTP).Experimental tests confirmed the results of the study on a mobile device.During the experiment,parameters such as the quality of recognition and tracking of moving objects,the PFTP and BAT,and the configuration parameters of the neural network and boundary line model were analyzed.The results showed an increased tracking accuracy with the proposed methods by 50%.The study results confirmed the relevance of the topic and the effectiveness of the implemented methods and tools.
基金supported by the National Natural Science Foundation of China(62303095)the Natural Science Foundation of Sichuan Province(2023NSFSC0872).
文摘Dear Editor,This letter proposes a novel Nash bargaining solution-based multiobjective model predictive control(MPC)scheme to deal with the interaction force control and the path-following problem of the constrained interactive robot.Considering the elastic interaction force model,a mechanical trade-off always exists between the interaction force and position,which means that neither force nor path following can satisfy their desired demands completely.Based on this consideration,two irreconcilable control specifications,the force object function and the position track object function,are proposed,and a new multi-objective MPC scheme is then designed.
基金supported by Priority Research Centers Program through NRF funded by MEST(2018R1A6A1A03024003)the Grand Information Technology Research Center support program IITP-2020-2020-0-01612 supervised by the IITP by MSIT,Korea.
文摘In the new era of technology,daily human activities are becoming more challenging in terms of monitoring complex scenes and backgrounds.To understand the scenes and activities from human life logs,human-object interaction(HOI)is important in terms of visual relationship detection and human pose estimation.Activities understanding and interaction recognition between human and object along with the pose estimation and interaction modeling have been explained.Some existing algorithms and feature extraction procedures are complicated including accurate detection of rare human postures,occluded regions,and unsatisfactory detection of objects,especially small-sized objects.The existing HOI detection techniques are instancecentric(object-based)where interaction is predicted between all the pairs.Such estimation depends on appearance features and spatial information.Therefore,we propose a novel approach to demonstrate that the appearance features alone are not sufficient to predict the HOI.Furthermore,we detect the human body parts by using the Gaussian Matric Model(GMM)followed by object detection using YOLO.We predict the interaction points which directly classify the interaction and pair them with densely predicted HOI vectors by using the interaction algorithm.The interactions are linked with the human and object to predict the actions.The experiments have been performed on two benchmark HOI datasets demonstrating the proposed approach.
基金Project supported by the National Natural Science Foundation of China(Grant Nos.31100672,51379121 and 61304230)the Shanghai Key Technology Plan Project(Grant Nos.12510501800,13510501600)
文摘There is a huge demand to develop a method for marine search and rescue（SAR） operators automatically predicting the most probable searching area of the drifting object. This paper presents a novel drifting prediction model to improve the accuracy of the drifting trajectory computation of the sea-surface objects. First, a new drifting kinetic model based on the geometry characteristics of the objects is proposed that involves the effects of the object shape and stochastic motion features in addition to the traditional factors of wind and currents. Then, a computer simulation-based method is employed to analyze the stochastic motion features of the drifting objects, which is applied to estimate the uncertainty parameters of the stochastic factors of the drifting objects. Finally, the accuracy of the model is evaluated by comparison with the flume experimental results. It is shown that the proposed method can be used for various shape objects in the drifting trajectory prediction and the maritime search and rescue decision-making system.
基金the National Natural Science Foundation of China(Grant Nos.61572264,61620106008)CAST young talents plan.
文摘Recent advances in supervised salient object detection modeling has resulted in significant performance improvements on benchmark datasets. However, most of the existing salient object detection models assume that at least one salient object exists in the input image. Such an assumption often leads to less appealing saliency maps on the background images with no salient object at all. Therefore, handling those cases can reduce the false positive rate of a model. In this paper, we propose a supervised learning approach for jointly addressing the salient object detection and existence prediction problems. Given a set of background-only images and images with salient objects, as well as their salient object annotations, we adopt the structural SVM framework and formulate the two problems jointly in a single integrated objective function: saliency labels of superpixels are involved in a classification term conditioned on the salient object existence variable, which in turn depends on both global image and regional saliency features and saliency labels assignments. The loss function also considers both image-level and regionlevel mis-classifications. Extensive evaluation on benchmark datasets validate the effectiveness of our proposed joint approach compared to the baseline and state-of-the-art models.
文摘This paper discusses smart body sensor objects （BSOs）, including their networking and internetworking. Smartness can be incorpo-rated into BSOs by embedding virtualization, predictive analytics, and proactive computing and communications capabilities. A few use cases including the relevant privacy and protocol requirements are also presented. General usage and deployment eti-quette along with the relevant regulatory implications are then discussed.
基金supported by the National Natural Science Foundation of China (Nos.61100045,61165013,61003142,60902023,and 61171096)the China Postdoctoral Science Foundation (Nos.20090461346,201104697)+3 种基金the Youth Foundation for Humanities and Social Sciences of Ministry of Education of China (No.10YJCZH117)the Fundamental Research Funds for the Central Universities (Nos.SWJTU09CX035,SWJTU11ZT08)Zhejiang Provincial Natural Science Foundation of China (Nos.Y1100589,Y1080123)the Natural Science Foundation of Ningbo,China (No.2011A610175)
文摘Emerging technologies of wireless and mobile communication enable people to accumulate a large volume of time-stamped locations,which appear in the form of a continuous moving object trajectory.How to accurately predict the uncertain mobility of objects becomes an important and challenging problem.Existing algorithms for trajectory prediction in moving objects databases mainly focus on identifying frequent trajectory patterns,and do not take account of the effect of essential dynamic environmental factors.In this study,a general schema for predicting uncertain trajectories of moving objects with dynamic environment awareness is presented,and the key techniques in trajectory prediction arc addressed in detail.In order to accurately predict the trajectories,a trajectory prediction algorithm based on continuous time Bayesian networks（CTBNs） is improved and applied,which takes dynamic environmental factors into full consideration.Experiments conducted on synthetic trajectory data verify the effectiveness of the improved algorithm,which also guarantees the time performance as well.
基金supported by the National Natural Science Foundation of China(No.51507141)Key research and development plan of Shaanxi Province(No.2018ZDCXL-GY-10-04)+1 种基金the National Key Research and Development Program of China(No.2016YFC0401409)the Shaanxi provincial education office fund(No.17JK0547).
文摘Interval prediction of wind power,which features the upper and lower limits of wind power at a given confidence level,plays a significant role in accurate prediction and stability of the power grid integrated with wind power.However,the conventional methods of interval prediction are commonly based on a hypothetic probability distribution function,which neglects the correlations among various variables,leading to the decrease of prediction accuracy.Therefore,we improve the multi-objective interval prediction based on the conditional copula function,through which we can fully utilize the correlations among variables to improve prediction accuracy without an assumed probability distribution function.We use the multi-objective optimization method of nondominated sorting genetic algorithm-II(NSGA-II)to obtain the optimal solution set.The particular best solution is weighted by the prediction interval average width(PIAW)and prediction interval coverage probability(PICP)to pick the optimized solution in practical examples.Finally,we apply the proposed method to three wind power plants in different cities in China as examples forvalidation and obtain higher prediction accuracy compared with other methods,i.e.,relevance vector machine(RVM),artificial neural network(ANN),and particle swarm optimization kernel extreme learning machine(PSO-KELM).These results demonstrate the superiority and practicability of this method in interval prediction of wind power.
基金supported by the National Natural Science Foundation of China(61071187)Key Laboratory Foundation for Underwater Test and Control Technology(9140c260201110c26)
文摘When deriving the Fourier diffraction theorem based on the first-order Born approximation,the difference between wave number of the scattering object and that of the surrounding medium is ignored,causing substantial errors in sound scattering prediction.This paper modifies the Born approximation by taking into account the amplitude and phase changes between the scattering object and the water due to the wave number difference.By changing the radius and center position of the sampling circle in the Fourier domain,accuracy of the predicted sound scattering is improved.With the modified Born approximation,the computed far-field directional pattern of the scattered sound from a circular cylinder is in good agreement with the rigorous solution.Numerical calculations for several objects with different shapes are used to show applicability and effectiveness of the proposed method.
基金supported by the National Natural Science Foundation of China[grant number 41901328 and 41974108]the Strategic Priority Research Program of the Chinese Academy of Sciences[grant number XDA19080101]the National Key Research and Development Program of China[grant number 2016YFB0501503 and 2016YFB0501502].
文摘Augmented virtual environments(AVE)combine real-time videos with 3D scenes in a Digital Earth System or 3D GIS to present dynamic information and a virtual scene simultaneously.AVE can provide solutions for continuous tracking of moving objects,camera scheduling,and path planning in the real world.This paper proposes a novel approach for 3D path prediction of moving objects in a video-augmented indoor virtual environment.The study includes 3D motion analysis of moving objects,multi-path prediction,hierarchical visualization,and path-based multi-camera scheduling.The results show that these methods can give a closed-loop process of 3D path prediction and continuous tracking of moving objects in an AVE.The path analysis algorithms proved accurate and time-efficient,costing less than 1.3 ms to get the optimal path.The experiment ran a 3D scene containing 295,000 triangles at around 35 frames per second on a laptop with 1 GB of graphics card memory,which means the performance of the proposed methods is good enough to maintain high rendering efficiency for a video-augmented indoor virtual scene.
文摘针对通用目标检测算法在检测小目标时存在错检和漏检等问题,提出了一种小目标检测算法IPH(Involution Prediction Head),将其运用在YOLOv4和YOLOv5的检测头部分,在VOC2007数据集上的实验结果表明,运用IPH后的YOLOv4小目标检测精度APs(AP for small objects)相比原始算法提升了1.1%,在YOLOv5上的APs更是提升了5.9%。经智能交通检测数据集进一步检验,IPH算法和去下采样能有效提升小目标检测精度,减少错检和漏检的情况。
基金Supported by the National Natural Science Foundation of China under Grant(62301330,62101346)the Guangdong Basic and Applied Basic Research Foundation(2024A1515010496,2022A1515110101)+1 种基金the Stable Support Plan for Shenzhen Higher Education Institutions(20231121103807001)the Guangdong Provincial Key Laboratory under(2023B1212060076).
文摘Background Co-salient object detection(Co-SOD)aims to identify and segment commonly salient objects in a set of related images.However,most current Co-SOD methods encounter issues with the inclusion of irrelevant information in the co-representation.These issues hamper their ability to locate co-salient objects and significantly restrict the accuracy of detection.Methods To address this issue,this study introduces a novel Co-SOD method with iterative purification and predictive optimization(IPPO)comprising a common salient purification module(CSPM),predictive optimizing module(POM),and diminishing mixed enhancement block(DMEB).Results These components are designed to explore noise-free joint representations,assist the model in enhancing the quality of the final prediction results,and significantly improve the performance of the Co-SOD algorithm.Furthermore,through a comprehensive evaluation of IPPO and state-of-the-art algorithms focusing on the roles of CSPM,POM,and DMEB,our experiments confirmed that these components are pivotal in enhancing the performance of the model,substantiating the significant advancements of our method over existing benchmarks.Experiments on several challenging benchmark co-saliency datasets demonstrate that the proposed IPPO achieves state-of-the-art performance.
文摘The determination and optimization of Automatic Train Operation(ATO) control strategy is one of the most critical technologies for urban rail train operation. The practical ATO optimal control strategy must consider many goals of the train operation, such as safety, accuracy, comfort, energy saving and so on. This paper designs a set of efficient and universal multi-objective control strategy. Firstly, based on the analysis of urban rail transit and its operating environment, the multi-objective optimization model considering all the indexes of train operation is established by using multi-objective optimization theory. Secondly, Non-dominated Sorting Genetic Algorithm II(NSGA-II) is used to solve the model, and the optimal speed curve of train running is generated.Finally, the intelligent controller is designed by the combination of fuzzy controller algorithm and the predictive control algorithm, which can control and optimize the train operation in real time. Then the robustness of the control system can ensure and the requirements for multi-objective in train operation can be satisfied.
基金Supported by National Basic Research Program of China（973 Program）（2013CB035500） National Natural Science Foundation of China（61233004,61221003,61074061）+1 种基金 International Cooperation Program of Shanghai Science and Technology Commission （12230709600） the Higher Education Research Fund for the Doctoral Program of China（20120073130006）
基金supported by the National Natural Science Foundation of China(12474189)National Key R&D Programof China(2021YFA1202300)+2 种基金Foundation of the President of China Academy of Engineering Physics(YZJJZQ2023016)Sichuan Provincial Distinguished Young Scholars Project(2025NSFJQ0022)National Natural Science Foundation of China(52394163,52371223,52101255,12192284).
文摘High-entropy alloys(HEAs)have emerged as promising candidates for catalyst applications due to their inherent compositional,structural,and site-level diversities,which enable highly tunable catalytic properties.However,these complexities pose grand challenges for traditional“trial-and-error”experimentation or computationally expensive“brute-force”ab initio calculations.Machine learning(ML)demonstrates great potential to address these challenges by establishing efficient,scalable mappings from composition,structure or site environment to HEA properties.Among these properties,adsorption energy,which quantifies the binding strength between catalytic intermediates and surface sites,is a crucial indicator of catalytic activity.This review provides a comprehensive overview ofML-driven strategies for adsorption energy prediction in the context of HEAs.Two primary strategies are introduced:“direct”prediction from unrelaxed structure and“iterative”prediction viaML potential-guided relaxation modeling.Both strategies can leverage handcrafted features or end-toend frameworks such as graph neural networks.We also discuss how pretrained models on largescale databases can extend to out-of-domain HEA systems.Beyond methodology,we address key challenges and future directions,including benchmarking ML strategies,developing HEA-specific datasets,pretraining and fine-tuning,integrating chained ML models,advancing multi-objective optimization,and bridgingMLpredictions with experimental validation.By critically evaluating existing strategies and highlighting emerging trends,this review underscores the critical role of ML in advancing adsorption energy predictions,offering a foundation for accelerating the discovery and optimization of HEA catalysts.
基金supported by the Advanced Materials-National Science and Technology Major Project(Grant No.2024ZD0607000)the National Natural Science Foundation of China(No.62475177)the Sichuan International Science and Technology Innovation Cooperation Project(No.2024YFHZ0328).
文摘Energetic Materials(EMs)play important roles in military,civilian and aerospace fields.Energy and stability are the two most important but contradictory properties in practical application,thus leading to difficult challenges in developing new EMswith high comprehensive performance.Motivated by the challenge,we exploit a de novo design framework targeting multiple objectives by integrating deep learning generator,machine learning prediction models,Pareto front optimization and quantum mechanics(QM)validation.First,heat of explosion(Q)and bond dissociation energy(BDE)are calculated by high-precisionQMfor 778 explosives experimentally reported.With the reliable dataset,RNN coupled with transfer learning is exploited to generate a new massive search space with 2×10^(5)potential energeticmolecules.Qand BDE prediction models with high accuracy are further developed by data augmentation and improvements in feature representation and model architectures,to quickly and accurately evaluate these new energeticmolecules.The modified 3D-GNN achieves an R^(2)=0.95 for the Q prediction,while the XGBoost coupled with the feature complementarity and PADRE data augmentation performs best for the BDE prediction(R2=0.98).To screen energetic compounds with trade-off energy and stability from the vast new molecule space,the predicted values and uncertainties are simultaneously considered,and Pareto front-based multi-objective screening is conducted by using 2D P[I]metric.QM calculation confirms the superior performance of the top 60 candidates to CL-20 in Q.25 promising energetic molecules with high energy and desired stability,as well as synthesis feasibility provide valuable candidates for experimental development.Also,the design strategy can be extended to other material fields.
基金supported by the National Key Research and Development Program of China(Grant Nos.2022YFB4602101)the National Science Fund for Distinguished Young Scholars(Grant No.52225103)+7 种基金the National Natural Science Foundation−Outstanding Youth Foundation(Grant No.52322102)the Funds for International Cooperation and Exchange of the National Natural Science Foundation of China(Grant No.W2412068)the Joint Funds of the National Natural Science Foundation of China(U2441262)the“Ten Thousand Talent Program for Prestigious Teachers”(Grant No.ZYZZ2023001)the National Natural Science Foundation of China(Grant Nos.12335017,52271003,52471002)the National Science Foundation for Young Scientists of China(Grant Nos.52201171,52201172,52401203)the Financial Support from the Fundamental Research Fund for the Central Universities of China(Grant Nos.FRF-TP-22-001C2,FRF-TP-22-005C2,FRF-TP-24-05C)the China Nuclear Power Technology Research Institute Co.Ltd.,and the China heavy-duty gas turbin technology Co.Ltd.under the project of J721.
文摘Low Young’s modulus and high yield strength are concurrently needed to meet the performance requirements of metallic implant materials.The single-objective performance-oriented alloy design strategies face challenges in effectively addressing the inherent conflict between Young’s modulus and yield strength.In this study,we developed a machine learning model for multi-objective synergistic optimization of modulus and yield strength,successfully enabling simultaneous prediction of Young’s modulus and yield strength in the Ti-Zr-Hf-Nb-Ta-Mo-Sn alloy system.The critical features influencing the modulus and strength of the alloys were systematically analyzed and identified.Moreover,a series of complex concentrated alloy(CCAs)with low Young’s modulus and high yield strength were successfully prepared based on this model.The newly developed alloys exhibited a stable single-phase BCC(body-centered-cubic)structure with Young’s modulus in the range of 40–50 GPa,yield strength of 600–915MPa,and elastic admissible strain of approximately 1.5%.The multi-objective machine learning model developed in this study can synergistically optimize low Young’s modulus and high yield strength in complex alloys,providing a novel approach for the design of advanced biomedical alloys.
基金supported by the National Key Technology R&D Program(Project Number 2016YFB1001402)the National Natural Science Foundation of China(Project Numbers61521002,61772298)+1 种基金Research Grant of Beijing Higher Institution Engineering Research CenterTsinghua–Tencent Joint Laboratory for Internet Innovation Technology.
文摘We propose a novel problem revolving around two tasks:(i)given a scene,recommend objects to insert,and(ii)given an object category,retrieve suitable background scenes.A bounding box for the inserted object is predicted in both tasks,which helps downstream applications such as semiautomated advertising and video composition.The major challenge lies in the fact that the target object is neither present nor localized in the input,and furthermore,available datasets only provide scenes with existing objects.To tackle this problem,we build an unsupervised algorithm based on object-level contexts,which explicitly models the joint probability distribution of object categories and bounding boxes using a Gaussian mixture model.Experiments on our own annotated test set demonstrate that our system outperforms existing baselines on all sub-tasks,and does so using a unified framework.Future extensions and applications are suggested.
基金The authors would acknowledge the research support from the Air Force Office of Scientific Research(AFOSR)FA9550-16-1-0184 and the Office of Naval Research(ONR)N00014-16-1-2729.Large amount of simulations of RSOs have been supported by the HPC cluster in School of Engineering,Rutgers University.
文摘In this paper,the recently developed machine learning(ML)approach to improve orbit prediction accuracy is systematically investigated using three ML algorithms,including support vector machine(SVM),artificial neural network(ANN),and Gaussian processes(GPs).In a simulation environment consisting of orbit propagation,measurement,estimation,and prediction processes,totally 12 resident space objects(RSOs)in solar-synchronous orbit(SSO),low Earth orbit(LEO),and medium Earth orbit(MEO)are simulated to compare the performance of three ML algorithms.The results in this paper show that ANN usually has the best approximation capability but is easiest to overfit data;SVM is the least likely to overfit but the performance usually cannot surpass ANN and GPs.Additionally,the ML approach with all the three algorithms is observed to be robust with respect to the measurement noise.
