CPS: TTP Option: Medium: Discovering and Resolving Anomalies in Smart Cities

CPS：TTP 选项：中：发现并解决智慧城市中的异常情况

• 批准号: 2038612
• 负责人: Srinivasa Narasimhan
• 金额: $ 120万
• 依托单位: Carnegie-Mellon University
• 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: 2020-09-01 至 2024-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2038612&HistoricalAwards=false
• 关键词: CPS; TTP; Option; Medium; Discovering

Understanding complex activity due to humans and vehicles in a large environment like a city neighborhood or even an entire city is one of the main goals of smart cities. The activities are heterogeneous, distributed, vary over time and mutually interact in many ways, making them hard to capture and understand and mitigate issues in a timely manner. While there has been tremendous progress in capturing aggregate statistics that helps in traffic and city management as well as personal planning and scheduling, much of this work ignores anomalous patterns. Examples include protests, erratic driving, near accidents, construction zone activity, and numerous others. Discovering and resolving anomalies is challenging for many reasons as they are complex and rare, depend on the context and depend on the spatial and temporal extent over which they are observed. There are potentially a large number of anomalies or anomalous patterns, so they are impossible to label and describe manually. The PIs will conduct research to address automatic discovery and resolution of anomalous patterns in smart city visual data. The PIs will leverage the large amount of visual data they have access to ranging from cameras at many intersections in Pittsburgh, around the Carnegie Mellon University neighborhood, cameras installed on public buses, and physical distribution networks in the city. The project will include the following four closely integrated research thrusts: (1) Extracting anomalies in the presence of noise due to visual processing algorithms, (2) Automatically discover anomalies at different spatial and temporal scales with intelligent coordinated and distributed planning, (3) discovering the relationship with anomalies and context, and (4) Resolving Anomalies through Hard and Soft Actuation using both automatic and human-in-the-loop methods. The work will enable the following applications: Safer and more efficient roads, monitoring the roadway infrastructure and roadside, maximizing the distribution services and informing decisions on health policy (including COVID-19). The project will be conducted in collaboration with several stakeholders - multiple infrastructure and traffic management startups and local city government - in a comprehensive transition to practice program designed to deploy the research in the real world.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.

在城市、社区甚至整个城市这样的大环境中，理解人和车辆的复杂活动是智慧城市的主要目标之一。这些活动是异质的、分布式的，随着时间的推移而变化，并以许多方式相互作用，使得它们很难及时捕获和理解并缓解问题。虽然在获取有助于交通和城市管理以及个人规划和调度的汇总统计数据方面取得了巨大进展，但这些工作中的大部分都忽略了异常模式。例子包括抗议、不稳定的驾驶、险些发生的事故、施工区的活动，以及许多其他的事情。发现和解决异常是具有挑战性的，原因有很多，因为它们复杂而罕见，取决于环境和观察到它们的空间和时间范围。可能存在大量的异常或异常模式，因此不可能手动对其进行标记和描述。PIS将进行研究，以解决智能城市视觉数据中异常模式的自动发现和解决问题。PI将利用他们可以访问的大量视觉数据，从匹兹堡许多十字路口的摄像头，卡内基梅隆大学附近的摄像头，安装在公共汽车上的摄像头，以及城市中的物流网络。该项目将包括以下四项紧密结合的研究任务：(1)在存在噪声的情况下提取由于视觉处理算法而产生的异常，(2)通过智能协调和分布式规划自动发现不同时空尺度的异常，(3)发现与异常和背景的关系，以及(4)使用自动和人在回路的方法通过硬执行和软执行解决异常。这项工作将使以下应用成为可能：更安全和更高效的道路，监测道路基础设施和路边，最大化分配服务，并为卫生政策决策提供信息(包括新冠肺炎)。该项目将与多个利益相关者--多家基础设施和交通管理初创公司以及当地市政府--合作进行，这是一个全面的实践过渡计划，旨在将研究部署到现实世界中。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势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

WALT: Watch And Learn 2D amodal representation from Time-lapse imagery

• DOI: 10.1109/cvpr52688.2022.00914
• 发表时间: 2022-06
• 期刊: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
• 作者: Dinesh Reddy Narapureddy;R. Tamburo;S. Narasimhan

Incorporating Queue Dynamics into Schedule-Driven Traffic Control

将队列动态纳入调度驱动的流量控制

• 发表时间: 2021
• 期刊: International Joint Conference on Artificial Intelligence
• 作者: Hu, Hsu-Chieh;Hawkes, Allen;Smith, Stephen F.
• 通讯作者: Smith, Stephen F.

Traffic4D: Single View Longitudinal 4D Reconstruction of Repetitious Activity using Self-Supervised Experts

Traffic4D：使用自我监督专家对重复活动进行单视图纵向 4D 重建

• 发表时间: 2021
• 期刊: IEEE Intelligent Vehicles Symposium
• 作者: Li, Fangyu;Reddy, N. Dinesh;Chen, Xudong;Narasimhan, Srinivasa G.
• 通讯作者: Narasimhan, Srinivasa G.

Exploiting & Refining Depth Distributions with Triangulation Light Curtains

剥削

• 发表时间: 2021
• 期刊: IEEE Conference on Computer Vision and Pattern Recognition
• 作者: Yaadhav, Raaj;Ancha, Siddharth;Tamburo, Robert;Held, David;Narasimhan, Srinivasa G.
• 通讯作者: Narasimhan, Srinivasa G.

Learning Continuous Implicit Representation for Near-Periodic Patterns

• DOI: 10.48550/arxiv.2208.12278
• 发表时间: 2022-08
• 期刊: ArXiv
• 作者: B. Chen;Tiancheng Zhi;M. Hebert;S. Narasimhan