基于Multitask⁃YOLO网络的卫星帆板ISAR图像快速分割

随着空间技术的飞速发展,空间态势感知能力需求不断增加。与传统光学传感器相比,逆合成孔径雷达(Inverse synthetic aperture radar,ISAR)具有全天候、远距离高分辨率成像的能力,且成像不受光照条件的影响。此外,空间态势感知系统需要对周围航天器进行准确的评估,因此对空间目标部件识别能力的需求日益迫切。本文提出了一种基于YOLOv5结构的Multitask⁃YOLO网络,用于卫星ISAR图像中卫星帆板的识别和分割。首先,本文添加了分割解耦头来实现网络的分割功能。然后用空间金字塔池快速算法(Spatial pyramid pooling fast,SPPF)和距离交并比算法(Distance intersection over union,DIoU)代替原有结构,避免图像失真,加快收敛速度。通过在通道中引入注意机制,提高了分割和识别的准确性。最后使用模拟卫星的ISAR图像进行实验。结果表明,所提出的Multitask⁃YOLO网络高效、准确地实现了部件的识别和分割。与其他的识别和分割网络相比,该网络的平均精度(mean Average precision,mAP)和平均交并比(mean Intersection over union,mIoU)提高了约5%。此外,该网络的运行速度高达16.4 GFLOP,优于传统的多任务网络的性能。

Constraints Separation Based Evolutionary Multitasking for Constrained Multi-Objective Optimization Problems

Constrained multi-objective optimization problems(CMOPs)generally contain multiple constraints,which not only form multiple discrete feasible regions but also reduce the size of optimal feasible regions,thus they propose serious challenges for solvers.Among all constraints,some constraints are highly correlated with optimal feasible regions;thus they can provide effective help to find feasible Pareto front.However,most of the existing constrained multi-objective evolutionary algorithms tackle constraints by regarding all constraints as a whole or directly ignoring all constraints,and do not consider judging the relations among constraints and do not utilize the information from promising single constraints.Therefore,this paper attempts to identify promising single constraints and utilize them to help solve CMOPs.To be specific,a CMOP is transformed into a multitasking optimization problem,where multiple auxiliary tasks are created to search for the Pareto fronts that only consider a single constraint respectively.Besides,an auxiliary task priority method is designed to identify and retain some high-related auxiliary tasks according to the information of relative positions and dominance relationships.Moreover,an improved tentative method is designed to find and transfer useful knowledge among tasks.Experimental results on three benchmark test suites and 11 realworld problems with different numbers of constraints show better or competitive performance of the proposed method when compared with eight state-of-the-art peer methods.

A deep multimodal fusion and multitasking trajectory prediction model for typhoon trajectory prediction to reduce flight scheduling cancellation

Natural events have had a significant impact on overall flight activity,and the aviation industry plays a vital role in helping society cope with the impact of these events.As one of the most impactful weather typhoon seasons appears and continues,airlines operating in threatened areas and passengers having travel plans during this time period will pay close attention to the development of tropical storms.This paper proposes a deep multimodal fusion and multitasking trajectory prediction model that can improve the reliability of typhoon trajectory prediction and reduce the quantity of flight scheduling cancellation.The deep multimodal fusion module is formed by deep fusion of the feature output by multiple submodal fusion modules,and the multitask generation module uses longitude and latitude as two related tasks for simultaneous prediction.With more dependable data accuracy,problems can be analysed rapidly and more efficiently,enabling better decision-making with a proactive versus reactive posture.When multiple modalities coexist,features can be extracted from them simultaneously to supplement each other’s information.An actual case study,the typhoon Lichma that swept China in 2019,has demonstrated that the algorithm can effectively reduce the number of unnecessary flight cancellations compared to existing flight scheduling and assist the new generation of flight scheduling systems under extreme weather.

RAIENet:End-to-End Multitasking Road All Information Extractor

Road lanes and markings are the bases for autonomous driving environment perception.In this paper,we propose an end-to-end multi-task network,Road All Information Extractor named RAIENet,which aims to extract the full information of the road surface including road lanes,road markings and their correspondences.Based on the prior knowledge of pavement information,we explore and use the deep progressive relationship between lane segmentation and pavement mark-ing detection.Then,different attention mechanisms are adapted for different tasks.A lane detection accuracy of 0.807 F1-score and a ground marking accuracy of 0.971 mean average precision at intersection over union(IOU)threshold 0.5 were achieved on the newly labeled see more on road plus(CeyMo+)dataset.Of course,we also validated it on two well-known datasets Berkeley Deep-Drive 100K(BDD100K)and CULane.In addition,a post-processing method for generating bird’s eye view lane(BEVLane)using lidar point cloud information is proposed,which is used for the construction of high-definition maps and subsequent decision-making planning.The code and data are available at https://github.com/mayberpf/RAIEnet.

Surface and Content Validity of an Advanced Beginner Nurse’s Self-Monitoring Scale While Multitasking

Background: Self-monitoring is important for recognizing the situations one is facing and assessing one’s own competence to respond appropriately to situations that require multitasking. Purpose: This study aimed to examine the surface and content validity of the Advanced Beginner Nurses’ Self-Monitoring Scale While Multitasking and refine the scale items accordingly. It is expected that the development of such scale will allow for reflection on advanced beginner nurses’ response to multitasking, leading to further capacity building. Methods: The surface validity of 96 items of the Advanced Beginner Nurses’ Self-Monitoring Scale While Multitasking was examined at a meeting with five expert researchers. Five researchers and five nurses examined the items’ content using an item-level content validity index through a questionnaire survey. Results and Conclusion: The Advanced Beginner Nurses’ Self-Monitoring Scale While Multitasking was organized into 73 items that were refined into scales with surface and content validity. Consequently, five sub-concepts were identified: recognizing the situation one’s facing, seeing one’s self from multiple perspectives, devising concrete strategies depending on the situation, considering a predictable time schedule, and being aware of the situation surrounding one’s self. In the future, it will be necessary to examine the reliability and validity of the scale.

Evolutionary Multitasking With Global and Local Auxiliary Tasks for Constrained Multi-Objective Optimization

Constrained multi-objective optimization problems(CMOPs) include the optimization of objective functions and the satisfaction of constraint conditions, which challenge the solvers.To solve CMOPs, constrained multi-objective evolutionary algorithms(CMOEAs) have been developed. However, most of them tend to converge into local areas due to the loss of diversity. Evolutionary multitasking(EMT) is new model of solving complex optimization problems, through the knowledge transfer between the source task and other related tasks. Inspired by EMT, this paper develops a new EMT-based CMOEA to solve CMOPs, in which the main task, a global auxiliary task, and a local auxiliary task are created and optimized by one specific population respectively. The main task focuses on finding the feasible Pareto front(PF), and global and local auxiliary tasks are used to respectively enhance global and local diversity. Moreover, the global auxiliary task is used to implement the global search by ignoring constraints, so as to help the population of the main task pass through infeasible obstacles. The local auxiliary task is used to provide local diversity around the population of the main task, so as to exploit promising regions. Through the knowledge transfer among the three tasks, the search ability of the population of the main task will be significantly improved. Compared with other state-of-the-art CMOEAs, the experimental results on three benchmark test suites demonstrate the superior or competitive performance of the proposed CMOEA.

Evolutionary Multitask Optimization in Real-World Applications: A Survey

Because of its strong ability to solve problems,evolutionary multitask optimization(EMTO)algorithms have been widely studied recently.Evolutionary algorithms have the advantage of fast searching for the optimal solution,but it is easy to fall into local optimum and difficult to generalize.Combining evolutionary multitask algorithms with evolutionary optimization algorithms can be an effective method for solving these problems.Through the implicit parallelism of tasks themselves and the knowledge transfer between tasks,more promising individual algorithms can be generated in the evolution process,which can jump out of the local optimum.How to better combine the two has also been studied more and more.This paper explores the existing evolutionary multitasking theory and improvement scheme in detail.Then,it summarizes the application of EMTO in different scenarios.Finally,according to the existing research,the future research trends and potential exploration directions are revealed.

BIM Supporting the Development of Multitasks Related with the Structural Project

Building Information Modelling (BIM) is a methodology focused on the centralization and sharing of the project information among all professionals involved, supported on the generation and manipulation of a three-dimensional (3D) digital BIM model. This methodology allows a close collaboration between the architect and the structural engineer and an adequate manipulation of the structural BIM model database, on the definition of multitasks. The collaboration allowed between all disciplines, avoid the detection of conflicts and data omission after in the construction place. Two BIM structural design cases were developed, using Revit as the modelling system and Robot as the structural software. Concerning the structural project the interoperability capacity between the software is still a limitation that engineers must be warned of. In the present study, the benefits and limitations identified within the communication and integration of distinct disciplines and on the development of most frequent multitasks normally related with a structural project, were considered.

Multitask Data Processing in a Wireless Alarm System

Aim To achieve multitask data procssing in a wireless alarm system by computer. Methods The alarm system was composed of hardware and software. The hardware was composed of a master master computer and slave transmitters. On urgent ugent occasion, one or more of the transmitters transmitted alarm signals and the master computer received the signals; interruption, residence, graph and word processing were utilized in software to achieve multitiask data processing . Results The main computer can conduct precise and quick multitask data procesing in any condition so long as alarm signals are received. The processing speed is higher than ordinary alarm System. Conclusion The master computer can conduct safe and quick multitask data processing by way of reliable design of software and hardware , so there is no need of special processor.

The Application of Multitasking Mechanism in Single Chip Computer System

Developed a new program structure using in single chip computer system, which based on multitasking mechanism. Discussed the specific method for realization of the new structure. The applied sample is also provided.

Performance Analysis of Robotic Arm Manipulators Control System under Multitasking Environment

This work is to observe the performance of PC based robot manipulator under general purpose （Windows）, Soft （Linux） and Hard （RT Linux） Real Time Operating Systems （OS）. The same open loop control system is observed in different operating systems with and without multitasking environment. The Data Acquisition （DAQ, PLC-812PG） card is used as a hardware interface. From the experiment, it could be seen that in the non real time operating system （Windows）, the delay of the control system is larger than the Soft Real Time OS （Linux）. Further, the authors observed the same control system under Hard Real Time OS （RT-Linux）. At this point, the experiment showed that the real time error （jitter） is minimum in RT-Linux OS than the both of the previous OS. It is because the RT-Linux OS kernel can set the priority level and the control system was given the highest priority. The same experiment was observed under multitasking environment and the comparison of delay was similar to the preceding evaluation.

Design and realization of a novel multitask TT&C operation pattern

With the sharp increase of China＇s in-orbit spacecraft and the constraint TT＆C resources, a mathematical model for optimal TT＆C resource allocation is proposed, and the TT＆C facility remote monitoring function is designed to achieve the multitask operation pattern under the unified management of the network management center. With this pattern, the TT＆C network management and the spacecraft management are separated, which is quite different from the previous pattern. Further, a novel spacecraft TT＆C technique based on spacecraft control language is developed, and the telecommanding pattern is designed to address the spacecraft operation problems. The engineering application shows that this pattern fundamentally improves the TT＆C network capability, increases the resource efficiency, and satisfies the efficient, accurate, and flexible operation of spacecraft.

An improved adaptive differential evolution algorithm for single unmanned aerial vehicle multitasking

Single unmanned aerial vehicle(UAV)multitasking plays an important role in multiple UAVs cooperative control,which is as well as the most complicated and hardest part.This paper establishes a threedimensional topographical map,and an improved adaptive differential evolution(IADE)algorithm is proposed for single UAV multitasking.As an optimized problem,the efficiency of using standard differential evolution to obtain the global optimal solution is very low to avoid this problem.Therefore,the algorithm adopts the mutation factor and crossover factor into dynamic adaptive functions,which makes the crossover factor and variation factor can be adjusted with the number of population iteration and individual fitness value,letting the algorithm exploration and development more reasonable.The experimental results implicate that the IADE algorithm has better performance,higher convergence and efficiency to solve the multitasking problem compared with other algorithms.

Optimal Design of a Ship Multitasking Cabin Layout Based on the Interval Optimization Method

Searching for the optimal cabin layout plan is an efective way to improve the efciency of the overall design and reduce a ship’s operation costs.The multitasking states of a ship involve several statuses when facing diferent missions during a voyage,such as the status of the marine supply and emergency escape.The human fow and logistics between cabins will change as the state changes.An ideal cabin layout plan,which is directly impacted by the above-mentioned factors,can meet the diferent requirements of several statuses to a higher degree.Inevitable deviations exist in the quantifcation of human fow and logistics.Moreover,uncontrollability is present in the fow situation during actual operations.The coupling of these deviations and uncontrollability shows typical uncertainties,which must be considered in the design process.Thus,it is important to integrate the demands of the human fow and logistics in multiple states into an uncertainty parameter scheme.This research considers the uncertainties of adjacent and circulating strengths obtained after quantifying the human fow and logistics.Interval numbers are used to integrate them,a two-layer nested system of interval optimization is introduced,and diferent optimization algorithms are substituted for solving calculations.The comparison and analysis of the calculation results with deterministic optimization show that the conclusions obtained can provide feasible guidance for cabin layout scheme.

Multitasking Behavior and Perceptions of Academic Performance in University Business Students in Mexico during the COVID-19 Pandemic

The current study measures the influence of multitasking behavior and self-efficacy for self-regulated learning(SESRL)on perceptions of academic performance and views in university students during the COVID-19 pan-demic in Mexico.264 university students fulfilled an online questionnaire.It was observed that multitasking beha-vior negatively influences SESRL(-0.203),while SESRL showed a positive influence of 0.537 on perceptions of academic performance,and multitasking behavior had an influence of-0.097 on the perception of academic per-formance.Cronbach’s alpha and Average Variance Extracted values were 0.809 and 0.577(multitasking behavior),0.819 and 0.626(SESRL),0.873 and 0.725(perceptions of academic performance),respectively.The results of the bootstrapping test showed that the path coefficients were significant.The study outcomes can support new plans in universities to ensure the best academic outcomes.Our study showed evidence of the COVID-19 impact on education behavior.This study’s novelty is based on using the partial least square structural equation modeling(PLS-SEM)technique to evaluate these variables.

From Rubbish to a Large Scale Industry： A Simple Fabrication of Superfiber with Multitasking Applications

In the whole earth, people increased dramatically from generation to generation which had created a large scale of broken environment so that people are facing more various types of garbage. Most of garbages are not useful and as a matter of fact, they are used to be neglected. Furthermore, many efforts have been conducted to change it by many types of recycled methods. Here, a simple technique is proposed with and without using fires to transform the useless natural or man-made rubbish things to be a superfiber as well as thin film with multitasking applications in human daily life. Since most of earth environment is covered by oceans, here the authors show how the ocean related garbage such as the crab skins, broken coral reefs and beach stones were changed to be superfiber and a multitasking device prototype.

MULTITASK SCHEDULING IN NETWORKED CONTROL SYSTEMS WITH APPLICATION TO LARGE SCALE VEHICLE CONTROL

Aiming at scheduling problems of networked control system （NCS） used to fulfill motion synthesis and cooperation control of the distributed multi-mechatronic systems, the differences of network scheduling and task scheduling are compared, and the mathematic description of task scheduling is presented. A performance index function of task scheduling of NCS according to task balance and traffic load matching principles is defined. According to this index, a static scheduling method is designed and implemented to controlling task set simulation of the DCY100 transportation vehicle. The simulation results are applied successfully to practical engineering in this case so as to validate the effectiveness of the proposed performance index and scheduling algorithm.

Multitask Learning with Multiscale Residual Attention for Brain Tumor Segmentation and Classification

Automatic segmentation and classification of brain tumors are of great importance to clinical treatment.However,they are challenging due to the varied and small morphology of the tumors.In this paper,we propose a multitask multiscale residual attention network(MMRAN)to simultaneously solve the problem of accurately segmenting and classifying brain tumors.The proposed MMRAN is based on U-Net,and a parallel branch is added at the end of the encoder as the classification network.First,we propose a novel multiscale residual attention module(MRAM)that can aggregate contextual features and combine channel attention and spatial attention better and add it to the shared parameter layer of MMRAN.Second,we propose a method of dynamic weight training that can improve model performance while minimizing the need for multiple experiments to determine the optimal weights for each task.Finally,prior knowledge of brain tumors is added to the postprocessing of segmented images to further improve the segmentation accuracy.We evaluated MMRAN on a brain tumor data set containing meningioma,glioma,and pituitary tumors.In terms of segmentation performance,our method achieves Dice,Hausdorff distance(HD),mean intersection over union(MIoU),and mean pixel accuracy(MPA)values of 80.03%,6.649 mm,84.38%,and 89.41%,respectively.In terms of classification performance,our method achieves accuracy,recall,precision,and F1-score of 89.87%,90.44%,88.56%,and 89.49%,respectively.Compared with other networks,MMRAN performs better in segmentation and classification,which significantly aids medical professionals in brain tumor management.The code and data set are available at https://github.com/linkenfaqiu/MMRAN.

Multitask Graph Neural Network for Knowledge Graph Link Prediction

Predicting entities in knowledge graphs is a crucial research area,and convolutional neural networks(CNNs)have exhibited significant performance due to their ability to generate expressive feature embeddings.However,sev-eral existing methods in thisfield tend to disrupt entities and relational embed-dings,disregarding the original translation characteristics in triples,leading to incomplete feature extraction.To address this issue and preserve the translation characteristics of triples,the present study introduces a novel representation tech-nique,termed MultiGNN.The suggested approach uses a graph convolutional neural network for encoding and implements a parameter sharing technique.It employs a convolutional neural network and a translation model as decoders.The model’s parameter space is expanded to effectively integrate translation charac-teristics into the convolutional neural network,which allows it to capture these characteristics and enhance the model’s performance.The proposed method in this paper has demonstrated significant enhancements in several metrics on the public benchmark dataset when compared to the baseline method.

面向合同信息抽取的动态多任务学习方法

对于合同文本中要素和条款两类信息的准确提取,可以有效提升合同的审查效率,为贸易各方提供便利化服务.然而当前的合同信息抽取方法一般训练单任务模型对要素和条款分别进行抽取,并没有深挖合同文本的特征,忽略了不同任务间的关联性.因此,采用深度神经网络结构对要素抽取和条款抽取两个任务间的相关性进行研究,并提出多任务学习方法.所提方法首先将上述两种任务进行融合,构建一种应用于合同信息抽取的基本多任务学习模型;然后对其进行优化,利用Attention机制进一步挖掘其相关性,形成基于Attention机制的动态多任务学习模型;最后针对篇章级合同文本中复杂的语义环境,在前两者的基础上提出一种融合词汇知识的动态多任务学习模型.实验结果表明,所提方法可以充分捕捉任务间的共享特征,不仅取得了比单任务模型更好的信息抽取结果,而且能够有效解决合同文本中要素与条款间实体嵌套的问题,实现合同要素与条款的信息联合抽取.此外,为了验证该方法的鲁棒性,在多个领域的公开数据集上进行实验,结果表明该方法的效果均优于基线方法.