Publications

7 Results

Search results

Jump to search filters

No tImplementing transition--edge sensors in a tabletop edge sensors in a tabletop xx--ray CT system for imaging applicationsray CT system for imaging applicationsitle

Alpert, Bradley; Becker, Daniel; Bennett, Douglas; Doriese, W.; Durkin, Malcolm; Fowler, Joseph; Gard, Johnathon; Imrek, Jozsef; Levine, Zachary; Mates, John; Miaja-Avila, Luis; Morgan, Kelsey; Nakamura, Nathan; O'Neil, Galen; Ortiz, Nathan; Reintsema, Carl; Schmidt, Daniel; Swetz, Daniel; Szypryt, Paul; Ullom, Joel; Vale, Leila; Weber, Joel; Wessels, Abigail; Dagel, Amber L.; Dalton, Gabriella D.; Laros, James H.; Jimenez, Edward S.; McArthur, Daniel M.; Thompson, Kyle R.; Walker, Christopher W.; Wheeler, Jason W.; Ablerto, Julien; Griveau, Damien; Silvent, Jeremie

Abstract not provided.

Phenomenology-informed techniques for machine learning with measured and synthetic SAR imagery

Proceedings of SPIE - The International Society for Optical Engineering

Walker, Christopher W.; Laros, James H.; Erteza, Ireena A.; Bray, Brian K.

Phenomenology-Informed (PI) Machine Learning is introduced to address the unique challenges faced when applying modern machine-learning object recognition techniques to the SAR domain. PI-ML includes a collection of data normalization and augmentation techniques inspired by successful SAR ATR algorithms designed to bridge the gap between simulated and real-world SAR data for use in training Convolutional Neural Networks (CNNs) that perform well in the low-noise, feature-dense space of camera-based imagery. The efficacy of PI-ML will be evaluated using ResNet, EfficientNet, and other networks, using both traditional training techniques and all-SAR transfer learning.

More Details

Phenomenology-informed techniques for machine learning with measured and synthetic SAR imagery

Proceedings of SPIE - The International Society for Optical Engineering

Walker, Christopher W.; Laros, James H.; Erteza, Ireena A.; Bray, Brian K.

Phenomenology-Informed (PI) Machine Learning is introduced to address the unique challenges faced when applying modern machine-learning object recognition techniques to the SAR domain. PI-ML includes a collection of data normalization and augmentation techniques inspired by successful SAR ATR algorithms designed to bridge the gap between simulated and real-world SAR data for use in training Convolutional Neural Networks (CNNs) that perform well in the low-noise, feature-dense space of camera-based imagery. The efficacy of PI-ML will be evaluated using ResNet, EfficientNet, and other networks, using both traditional training techniques and all-SAR transfer learning.

More Details
7 Results
7 Results