MRI: Acquisition of Omnipercipient Chamber for Gathering Ground Truth and Enabling Research on Smart and Connected Things

MRI：获取全知室以收集地面真相并支持智能和互联事物的研究

• 批准号: 2018966
• 负责人: Srihari Nelakuditi
• 金额: $ 50万
• 依托单位: University of South Carolina at Columbia
• 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: 2020-10-01 至 2024-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2018966&HistoricalAwards=false
• 关键词: MRI; Acquisition; Omnipercipient; Chamber; Gathering

This project supports the acquisition of an omnipercipient chamber (OmniC), with all its sides mounted with a variety of antennas and sensors, for gathering the ground truth (i.e., the information provided by direct observation as opposed to indirect inference) on experiments conducted inside it. This instrument enables the research in diverse areas such as wireless networking, therapeutic and rehabilitation care, robotics and computer vision, cyber physical systems, and augmented and virtual reality. OmniC addresses the needs of many researchers, reduces the cost and effort of pursuing research, furthers inter-disciplinary collaborations, and consequently improves the speed and accuracy of scientific advancement. The labeled ground truth data generated from the experiments conducted inside OmniC would be widely disseminated to help the global research community. This project will also broaden the participation by involving women, minorities, and undergraduates in research, with the help of women investigators and local HBCUs. In addition to assimilating the research results into graduate and undergraduate level courses, this project will showcase the research advances to a wider audience in coordination with the University of South Carolina's outreach activities.The core benefit of this instrument is that it captures the phenomenon being observed coherently across the time, space, and frequency domains. The research activities enabled by OmniC include: i) Devising next-generation mmWave and massive MIMO wireless networks for high throughput and reliability by utilizing channel sparsity; ii) Designing reconfigurable antennas that operate over a broad spectrum from microwave to mmWave; iii) Characterizing thermal dissipation of mmWave devices and developing antenna scheduling schemes for adaptive cooling; iv) Assessing mobility impairments of patients and elderly leveraging the wireless signal reflections and floor vibrations for gait measurement; v) Coordinating autonomous surface vehicles and autonomous aerial vehicles during takeoff and landing under inclement weather and communication constraints; vi) Tracking the movements and poses of subjects across multiple videos and recognizing their emotions by building and training deep learning models; vii) Canceling acoustic noise with IoT relays exploiting the difference in propagation speeds of sound and wireless signals; viii) Supporting ultra-low latency virtual and augmented reality applications over mmWave networks through fast beam alignment protocols. Overall, this instrument taps the common needs and synergy between a variety of research groups across disciplines in advancing research on smart and connected things.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.

该项目支持获得一个全感知室（OmniC），其所有侧面都安装了各种天线和传感器，用于收集在其中进行的实验的地面真相（即，由直接观察提供的信息，而不是间接推断）。该仪器使研究在不同的领域，如无线网络，治疗和康复护理，机器人和计算机视觉，网络物理系统，增强和虚拟现实。OmniC解决了许多研究人员的需求，降低了进行研究的成本和努力，进一步促进了跨学科合作，从而提高了科学进步的速度和准确性。在OmniC内部进行的实验中产生的标记的地面真相数据将被广泛传播，以帮助全球研究界。该项目还将在女性调查人员和当地hbcu的帮助下，扩大妇女、少数民族和本科生参与研究的范围。除了将研究成果融入研究生和本科水平的课程之外，该项目还将与南卡罗来纳大学的外展活动协调，向更广泛的受众展示研究进展。该仪器的核心优势在于，它能够捕捉到在时间、空间和频域上一致观察到的现象。OmniC支持的研究活动包括：i)利用信道稀疏性设计下一代毫米波和大规模MIMO无线网络，实现高吞吐量和可靠性；ii)设计可重构天线，在从微波到毫米波的广谱范围内工作；iii)研究毫米波器件的热耗散特性，制定自适应冷却的天线调度方案；iv)利用无线信号反射和地板振动测量步态，评估患者和老年人的行动障碍；v)在恶劣天气和通信条件下，协调自主地面车辆和自主飞行器的起降；vi)通过建立和训练深度学习模型，跟踪多个视频中受试者的动作和姿势，并识别他们的情绪；vii)利用物联网中继利用声音和无线信号传播速度的差异来消除噪声；viii)通过快速波束对准协议，在毫米波网络上支持超低延迟虚拟和增强现实应用。总体而言，该工具利用了不同学科研究小组之间的共同需求和协同作用，以推进智能和互联事物的研究。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识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

Towards Deep Learning Augmented Robust D-Band Millimeter-Wave Picocell Deployment

迈向深度学习增强稳健 D 频段毫米波微微蜂窝部署

• DOI: 10.1145/3595244.3595266
• 发表时间: 2023
• 期刊: ACM SIGMETRICS Performance Evaluation Review
• 作者: Regmi, Hem;Sur, Sanjib
• 通讯作者: Sur, Sanjib

Is It Necessary to Transfer Temporal Knowledge for Domain Adaptive Video Semantic Segmentation?

• DOI: 10.1007/978-3-031-19812-0_21
• 发表时间: 2022
• 作者: Xinyi Wu;Zhenyao Wu;J. Wan;Lili Ju;Song Wang

A Case for Line-Of-Sight Blockage Detection as a Primitive in Millimeter-Wave Networks

视距阻塞检测作为毫米波网络原语的案例

• 发表时间: 2022
• 期刊: IEEE Internatonal Conference on Mobile Adhoc and Sensor Systems (MASS
• 作者: Sur, S;Nelakuditi, S
• 通讯作者: Nelakuditi, S

SiamDoGe: Domain Generalizable Semantic Segmentation Using Siamese Network

SiamDoGe：使用 Siamese 网络进行领域可泛化语义分割

• 发表时间: 2022
• 期刊: Springer ECCV 202
• 作者: Zhang, X;Ju, Li
• 通讯作者: Ju, Li

SugarWave: A Non-destructive Estimation of Fruit Sugar Content Using Millimeter-Wave Sensing

SugarWave：利用毫米波传感无损估算水果糖含量

• DOI: 10.1109/mass58611.2023.00079
• 发表时间: 2023
• 期刊: IEEE
• 作者: Tavasoli, Reza;Sur, Sanjib;Nelakuditi, Srihari
• 通讯作者: Nelakuditi, Srihari