MRI: Acquisition of a GPU-Based Cloud Infrastructure for Inter-/Multi-Disciplinary Research and Education at a Primarily Undergraduate Institution

MRI：采购基于 GPU 的云基础设施，用于主要本科机构的跨/多学科研究和教育

• 批准号: 1726017
• 负责人: Shaoen Wu
• 金额: $ 26万
• 依托单位: Ball State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-10-01 至 2020-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1726017&HistoricalAwards=false
• 关键词: MRI; Acquisition; GPU; Based; Cloud

This project, acquiring equipment to establish a GPU-based cloud, aims to enable big data related interdisciplinary projects, specifically collaborative research projects in computer science, large-scale medical data, computational chemistry, and geology. These include:- Intelligent Internet of Things (IoT),- Smart Health,- Spatial Computing, and - Computational ChemistryThe intelligent IOT research would utilize the infrastructure to develop new deep learning models to enable smart living and navigation for vision impaired people. The big data enriched health research expects to find effective solutions to solve problems of obesity and non-communicable chronic diseases, serious threats to public health in the US and globally, using the instrumentation to compute the models with enormous health data. The geoscience research would enabled the establishment of "American Spatial Data Portal" (ASDP) that will provide a central repository of spatial data to expedite research on geospatial data mining and applications. A cloud geospatial education system would be developed on the instrument to transform the nation's geoscience education. The computational chemistry research would take advantage of the computation capability of the cloud to discover, design, and evaluate new and low costs catalysts for important chemical reactions, which could significantly change the life of the world. This GPU-based cloud is expected to facilitate the data and computation intensive research progress across many disciplines not only at this mainly undergraduate university, but also at other institutions through authorized and scheduled remote access. Furthermore, the infrastructure would be shared via high speed internet to other colleagues in the nation, and to international collaborators who are interested. Broader Impacts: By involving students in the activities of installation, configuration, and instrumentation, this project would provide valuable opportunities to train future scientists, engineers, and instrumentalists. The developed new cross-list courses and summer training program would train hundreds of both senior and graduate students on GPU and big data related knowledge and skills. Moreover, due to the appropriate mentoring that the investigators are able to offer, overall students and minorities can benefit.

该项目购买设备以建立基于 GPU 的云，旨在支持大数据相关的跨学科项目，特别是计算机科学、大规模医疗数据、计算化学和地质学方面的合作研究项目。其中包括：- 智能物联网 (IoT)、- 智能健康、- 空间计算和 - 计算化学智能物联网研究将利用基础设施开发新的深度学习模型，为视力障碍人士提供智能生活和导航。大数据丰富的健康研究期望找到有效的解决方案来解决肥胖和非传染性慢性疾病问题，这些问题对美国和全球的公共健康构成严重威胁，利用仪器通过大量健康数据计算模型。地球科学研究将有助于建立“美国空间数据门户”（ASDP），该门户将提供空间数据的中央存储库，以加快地理空间数据挖掘和应用的研究。将在该仪器上开发云地理空间教育系统，以改变国家的地球科学教育。计算化学研究将利用云的计算能力来发现、设计和评估用于重要化学反应的新型低成本催化剂，这可能会显着改变世界的生活。这种基于 GPU 的云预计将不仅在这所主要本科大学中，而且通过授权和预定的远程访问，促进跨多个学科的数据和计算密集型研究进展。此外，基础设施将通过高速互联网与国内其他同事以及感兴趣的国际合作者共享。更广泛的影响：通过让学生参与安装、配置和仪器仪表活动，该项目将为培训未来的科学家、工程师和乐器演奏家提供宝贵的机会。新开发的跨列表课程和暑期培训计划将为数百名高年级学生和研究生提供GPU和大数据相关知识和技能的培训。此外，由于研究人员能够提供适当的指导，全体学生和少数群体都可以受益。

项目成果

Indoor Multi-Sensory Self-Supervised Autonomous Mobile Robotic Navigation

室内多感官自监督自主移动机器人导航

• DOI: 10.1109/icii.2018.00021
• 发表时间: 2018
• 期刊: International Conference on Industrial Internet of Things 2018
• 作者: Xu, Juhong;Guo, Hanqing;Wu, Shaoen
• 通讯作者: Wu, Shaoen

Non-Contact Non-Invasive Heart and Respiration Rates Monitoring with MIMO Radar Sensing

利用 MIMO 雷达感应进行非接触式非侵入式心脏和呼吸频率监测

• DOI: 10.1109/glocom.2018.8648106
• 发表时间: 2018
• 期刊: IEEE Globecom 2018
• 作者: Liu, Qiwei;Guo, Hanqing;Xu, Junhong;Wang, Honggang;Kageza, Aron;AlQarni, Saeed;Wu, Shaoen
• 通讯作者: Wu, Shaoen

DSIC: Deep Learning Based Self-Interference Cancellation for In-Band Full Duplex Wireless

• DOI: 10.1109/globecom38437.2019.9013521
• 发表时间: 2018-11
• 期刊: 2019 IEEE Global Communications Conference (GLOBECOM)
• 作者: Hanqing Guo;N. Zhang;Saeed AlQarni;Shaoen Wu

Automated Labeling for Robotic Autonomous Navigation Through Multi-Sensory Semi-Supervised Learning on Big Data

通过大数据多感官半监督学习实现机器人自主导航的自动标记

• DOI: 10.1109/tbdata.2019.2892462
• 发表时间: 2019
• 期刊: IEEE Transactions on Big Data
• 影响因子: 7.2
• 作者: Xu, Junhong;Zhu, Shangyue;Guo, Hanqing;Wu, Shaoen
• 通讯作者: Wu, Shaoen

Shared Multi-Task Imitation Learning for Indoor Self-Navigation

• DOI: 10.1109/glocom.2018.8647614
• 发表时间: 2018-08
• 期刊: 2018 IEEE Global Communications Conference (GLOBECOM)
• 作者: Junhong Xu;Qiwei Liu;Hanqing Guo;Aaron Kageza;Saeed AlQarni;Shaoen Wu