Transfer learning for radioactive particle tracking
Published in Chemical Engineering Science, 2022
This paper investigates the use of transfer learning to improve calibration efficiency in radioactive particle tracking (RPT) for multiphase flow systems. We show that leveraging historical calibration data can significantly improve tracking accuracy when new calibration data are limited or unavailable. The results demonstrate that transfer learning enables reliable RPT modeling across changing experimental conditions, reducing the need for costly recalibration.
Recommended citation: Guilherme Lindner, Sai Shi, Slobodan Vučetić, Sanja Mišković. (2022). “Transfer learning for radioactive particle tracking” Chemical Engineering Science. Volume 248, Part B, 117190, ISSN 0009-2509.