Details
- Identification
- ISSN: 1977-5296, DOI: 10.3011/ESARDA.IJNSNP.2023.5
- Publication date
- 13 December 2023
- Author
- Joint Research Centre
Description
Volume: 65, December 2023, pages 34-43
Authors: Zoe N. Gastelum a, Timothy M. Shead b and Matthew Marshall c
a Sandia National Laboratories, International Safeguards & Engagements Department, 1515 Eubank SE, Albuquerque, NM, USA, 87123
b Sandia National Laboratories, Machine Intelligence and Visualization Department, 1515 Eubank SE, Albuquerque, NM, USA, 87123
c Sandia National Laboratories, Nuclear Verification Department, 1515 Eubank SE, Albuquerque, NM, USA, 87123
Abstract: Computer vision models have great potential as tools for international nuclear safeguards verification activities, but off-the-shelf models require fine-tuning through transfer learning to detect relevant objects. Because open-source examples of safeguards-relevant objects are rare, and to evaluate the potential of synthetic training data for computer vision, we present the Limbo dataset. Limbo includes both real and computer-generated images of uranium hexafluoride containers for training computer vision models. We generated these images iteratively based on results from data validation experiments that are detailed here. The findings from these experiments are applicable both for the safeguards community and the broader community of computer vision research using synthetic data.
Keywords: Computer vision, synthetic data, international nuclear safeguards, uranium hexafluoride.
Reference guideline:
Gastelum, Z.N., Shead, T.M., and Marshall, M., (2023, December). Data Validation Experiments with a Computer-Generated Imagery Dataset for International Nuclear Safeguards, ESARDA Bulletin - The International Journal of Nuclear Safeguards and Non-proliferation, 65, 44-62. https://doi.org/10.3011/ESARDA.IJNSNP.2023.5