Two-dimensional total absorption spectroscopy with conditional generative adversarial networks
share
We explore the use of machine learning techniques to remove the response of large volume γ-ray detectors from experimental spectra. Segmented γ-ray total absorption spectrometers (TAS) allow for the simultaneous measurement of individual γ-ray energy (E_γ) and total excitation energy (E_x). Analysis of TAS detector data is complicated by the fact that the E_x and E_γ quantities are correlated, and therefore, techniques that simply unfold using E_x and E_γ response functions independently are not as accurate. In this work, we investigate the use of conditional generative adversarial networks (cGANs) to simultaneously unfold E_x and E_γ data in TAS detectors. Specifically, we employ a Pix2Pix cGAN, a generative modeling technique based on recent advances in deep learning, to treat (E_x, E_γ) matrix unfolding as an image-to-image translation problem. We present results for simulated and experimental matrices of single-γ and double-γ decay cascades. Our model demonstrates characterization capabilities within detector resolution limits for upwards of 90% of simulated test cases.
READ FULL TEXT