CyberTraining: Implementation: Small: Collaborative and Integrated Training on Connected and Autonomous Vehicles Cyber Infrastructure

网络培训：实施：小型：联网和自动驾驶车辆网络基础设施的协作和综合培训

• 批准号: 2017564
• 负责人: Song Fu
• 金额: $ 49.94万
• 依托单位: University of North Texas
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-08-01 至 2024-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2017564&HistoricalAwards=false
• 关键词: CyberTraining; Implementation; Small; Collaborative; Integrated

In response to the quick development and adoption of connected and autonomous vehicles (CAVs), U.S. cities and states have recently started building CAV cyberinfrastructure. However, there is not an adequate supply of skilled research workforce in this field. A key obstacle to such workforce development is the lack of well-structured training programs for utilizing CAV cyberinfrastructure to enable and potentially transform fundamental CAV research. To address this issue, a project-oriented training program will be developed in this CyberTraining project to enable scientific research workforce development for CAV cyberinfrastructure. It is estimated that the adoption of CAVs would lead to nearly $800 billion in annual social and economic benefits by 2050, therefore, it is important for the nation to invest in CAV cyberinfrastructure research workforce training programs. The proposed training program targets students and early-stage researchers who are interested in CAVs, including participants with a broad diversity in academic level and in experience level with CAVs. It is expected that more than 100 trainees will participate every year in the training program, including researchers from various domains such as cyber-physical systems, edge computing, wireless networking, deep learning, computer vision, and big data. A longstanding collaboration with the trainees and/or their advisors will be built to ensure a broad adoption of CAV cyberinfrastructure by the research community to catalyze major research advances. The long-term goal of this project is to develop a first of its kind open CAV cyberinfrastructure, an integrated training and research hub, to accelerate research and education in CAVs.The goal of this project is to develop a collaborative and integrated training program to enable scientific research work force development for Connected and Autonomous Vehicle CyberInfrastructure (CAV-CI) and foster broad adoption of CAV-CI to advance fundamental CAV related research. To achieve these goals, the project will leverage existing partnerships with relevant stakeholders to create tailored, high-impact, engaging, collaborative, and integrated training modules for CAV-CI research workforce development. With the aim of enhancing trainees design and implementation capabilities, problem-solving skills, and critical thinking ability, the proposed training program will result in: (1) a project-oriented short course plus long-term coaching and support, (2) hands-on training modules on the perception, network, and application layers in CAV-CI, (3) an annual research workshop that disseminates research results and receives feedback on the training program from the research and industrial communities, and (4) research projects for students supported through NSF's Research Experiences for Undergraduates (REU)program and capstone projects for senior undergraduates. During the training workshops, project-oriented training will be offered to actively engage trainees in learning and solving real-world problems. Three sample research projects will be designed, allowing trainees to develop complete research skills, i.e., competency to solve authentic problems. Following every sample research project, with each having a strong practical relevance and meaningfulness, two versions of training modules will be developed to reach a broader trainee group: a fundamental training module for undergraduate students and community college educators, and a research-intensive training module for graduate students and postdocs. By taking either the fundamental or the research-intensive training modules, trainees will enhance their problem-solving skills, improve their creative and independent thinking ability, as well as gaining enthusiasm and confidence in conducting CAV-CI enabled research. The CAV-CI education, research and training activities include specific goals to train individuals from underrepresented groups and the broader STEM workforce.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

为了应对互联和自动驾驶汽车（CAV）的快速发展和采用，美国城市和州最近开始建设CAV网络基础设施。然而，在这一领域没有足够的熟练研究人员。这种劳动力发展的一个关键障碍是缺乏结构良好的培训计划，利用CAV网络基础设施，使和潜在地改变基本的CAV研究。为了解决这个问题，将在这个网络培训项目中开发一个面向项目的培训计划，以使CAV网络基础设施的科学研究队伍的发展。据估计，到2050年，CAV的采用将带来近8000亿美元的年度社会和经济效益，因此，国家投资CAV网络基础设施研究人员培训计划非常重要。拟议的培训计划针对对CAV感兴趣的学生和早期研究人员，包括学术水平和CAV经验水平广泛多样的参与者。预计每年将有100多名学员参加培训计划，其中包括来自网络物理系统、边缘计算、无线网络、深度学习、计算机视觉和大数据等各个领域的研究人员。将与学员和/或他们的顾问建立长期合作，以确保研究界广泛采用CAV网络基础设施，以促进重大研究进展。该项目的长期目标是开发第一个开放式CAV网络基础设施，一个综合培训和研究中心，加速CAV的研究和教育。该项目的目标是开发一个协作和综合培训计划，以促进互联和自动驾驶汽车网络基础设施（CAV-CI）的科学研究工作队伍的发展，并促进CAV的广泛采用。CI推进CAV相关基础研究。为了实现这些目标，该项目将利用与相关利益相关者的现有合作伙伴关系，为CAV-CI研究队伍的发展创建量身定制的，高影响力的，参与的，协作的和综合的培训模块。为了提高学员的设计和实施能力，解决问题的能力和批判性思维能力，拟议的培训计划将导致：（1）以项目为导向的短期课程加上长期的指导和支持，（2）CAV-CI中感知层、网络层和应用层的实践培训模块，（3）每年举办一次研究讲习班，传播研究成果，并听取研究界和工业界对培训方案的反馈意见，以及（4）通过NSF的本科生研究经验（REU）计划和高年级本科生的顶点项目支持的学生研究项目。在培训讲习班期间，将提供以项目为导向的培训，使学员积极参与学习和解决现实世界的问题。将设计三个样本研究项目，让学员发展完整的研究技能，即，解决真实问题的能力。在每个样本研究项目之后，每个项目都具有很强的实际相关性和意义，将开发两个版本的培训模块以覆盖更广泛的受训者群体：本科生和社区学院教育工作者的基础培训模块，以及研究生和博士后的研究密集型培训模块。通过参加基础或研究密集型培训模块，学员将提高他们解决问题的能力，提高他们的创造力和独立思考能力，并获得进行CAV-CI支持研究的热情和信心。 CAV-CI的教育、研究和培训活动包括培训来自代表性不足的群体和更广泛的STEM劳动力的个人的具体目标。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Learning Connected Attentions for Convolutional Neural Networks

• DOI: 10.1109/icme51207.2021.9428397
• 发表时间: 2021-07
• 期刊: 2021 IEEE International Conference on Multimedia and Expo (ICME)
• 作者: Xu Ma;Jingda Guo;Sihai Tang;Zhinan Qiao;Qi Chen;Qing Yang;Song Fu

LiDAR-based Cooperative Relative Localization

基于LiDAR的协同相对定位

• DOI: 10.1109/iv55152.2023.10186549
• 发表时间: 2023
• 期刊: 2023 IEEE Intelligent Vehicles Symposium (IV
• 作者: Dong, Jiqian;Chen, Qi;Qu, Deyuan;Lu, Hongsheng;Ganlath, Akila;Yang, Qing;Chen, Sikai;Labi, Samuel
• 通讯作者: Labi, Samuel

VECFrame: A Vehicular Edge Computing Framework for Connected Autonomous Vehicles

VECFrame：用于联网自动驾驶车辆的车辆边缘计算框架

• DOI: 10.1109/edge53862.2021.00019
• 发表时间: 2021
• 期刊: 2021 IEEE International Conference on Edge Computing (EDGE
• 作者: Tang, Sihai;Chen, Bruce;Iwen, Harold;Hirsch, Jason;Fu, Song;Yang, Qing;Palacharla, Paparao;Wang, Nannan;Wang, Xi;Shi, Weisong
• 通讯作者: Shi, Weisong

Spatial Pyramid Attention for Deep Convolutional Neural Networks

深度卷积神经网络的空间金字塔注意力

• DOI: 10.1109/tmm.2021.3068576
• 发表时间: 2021
• 期刊: IEEE Transactions on Multimedia
• 影响因子: 7.3
• 作者: Ma, Xu;Guo, Jingda;Sansom, Andrew;Mcguire, Mara;Kalaani, Andrew;Chen, Qi;Tang, Sihai;Yang, Qing;Fu, Song
• 通讯作者: Fu, Song

Machine-Learning-Enabled Cooperative Perception for Connected Autonomous Vehicles: Challenges and Opportunities

• DOI: 10.1109/mnet.011.2000560
• 发表时间: 2021-05
• 期刊: IEEE Network
• 影响因子: 9.3
• 作者: Qing Yang;Song Fu;Honggang Wang;Hua Fang