Deep Learning (DL) has been a concept since the late 1950s and has advanced significantly in the last twenty years due to enhanced computational architectures. With the increase in computer processing and graphical power, the utility of DL has moved from a concept to a highly praised capability across many industries. With the growth of DL in industry and research, the International Atomic Energy Agency (IAEA) has started to aggressively research the use of DL to enhance the ability in collection and analysis of international safeguard-relevant information. The ways industry has successful utilized DL can give insight into potential applications for international safeguards data analysis. In this paper, we will review the use of DL across multiple industries. The paper organized into four major sections: 1) the impact DL has on the current work force, 2) the traits of a dataset that make it a better candidate for DL models, 3) the different ways industries are using DL, and 4) potential transfers of DL capabilities to multiple safeguards activities. By understanding current applications of DL across multiple industries, the international safeguards research community can start to hone in on realistic potential applications in our domain and learn from both successes and failures from other domains.
Year
2020
Abstract