Advanced AI-Powered 3D Sensing: Fast Speed, High Accuracy, Full Field-of-View, and Broad Range Capabilities
Abstract
Recent a few years have seen an explosion of interest in the research and development of artificial intelligence (AI) techniques. In the meantime, the global economy has started benefiting drastically from the wide applications of AI technologies, which have taken over challenging job duties from humans to yield lower cost, higher safety, higher reliability, and greater productivity. The recently released OpenAI GPT-4 and other similar AI models have shown that AI could have an impact equivalent to that of the Internet, fueling the worldwide technology revolution in many aspects. AI technologies have become increasingly important to the US national defense due to their significant advantages in data processing, decision-making, situational awareness, operational efficiency, cybersecurity, hardware, weaponry, training, simulation, etc. As a result, many government agencies, including the Department of Defense (DoD), are actively supporting AI research and development in a variety of ways. Meanwhile, the principal investigator (PI) of this proposal has been dedicated to developing AI-based new techniques to cope with the limitations of existing technologies. These research tasks rely highly on a high-performance computer system suitable for training sophisticated AI and deep-leaning models. The proposed NVIDIA H100 system is designed to be the most advanced flexible hardware and software platform for AI researchers. The system will support much-needed significant research and education on fast-speed, high-accuracy, full-field, and board-range 3D sensing for numerous military and civil applications. In addition, it is noteworthy that a large number of researchers at the Catholic University of America (CUA) are performing research and educational activities that require a high-performance computer system; however, CUA does not have such an advanced system. This has become somewhat of a hindrance to the research and education productivity in the relevant areas at CUA. This project, if funded, will bring the very first high-performance computer with highly-demanded functionalities to CUA. The plentiful advanced features of the computer platform and system enable this versatile piece of equipment to be readily available for use by researchers, educators, and students at the university. The CUA School of Engineering (SoE) where the PI works has a number of off-campus graduate educational programs at DoD-affiliated agencies and centers. Particularly, the PI regularly teaches advanced courses at the Naval Surface Warfare Center (NSWC) at Carderock and the Army?s Night Vision and Electronic Sensors Directorate (NVESD) at Fort Belvoir. Furthermore, it has been a tradition that many students join DoD-affiliated divisions or DoD-related companies after graduation, and recently more faculty members working on DoD-relevant research are being recruited. Consequently, the NVIDIA H100 system will remarkably enhance the research infrastructure and substantially support the DoD-prioritized research and education activities of many faculty members and students at CUA. Despite CUA?s comparatively smaller size and research capabilities when compared to top engineering schools, we firmly believe that CUA has the potential to lead cutting-edge research and education work in the emerging field of AI and will continue to make notable contributions to the national defense. In light of this, we earnestly seek DoD?s support for our proposed system.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 16, 2023
- Source ID
- W911NF2310367
Entities
People
- Zhaoyang Wang
Organizations
- Army Contracting Command
- The Catholic University of America
- United States Army