Speaker
Description
High-purity germanium detectors are used in the search for rare events such as neutrinoless
double-beta decay, dark matter and other beyond Standard Model physic. Due to the
infrequent occurrence of signal events, extraordinary measures are taken to reduce background
interactons and extract the most informatio from data. An efficiensignal denoising algorithm
can improve energy resolutio and background rejectio techniques, and help classify signal
events. It can also help identify lo-energy events where the signal-to-noise ratio is smal.
In this work, we demonstrate the applicatio of generative adversarial networks withdeep
convolutional autoencodes to remove electronic noise from high-purity germanium p-type
point contact detector signals. Built on the success of denoising using a convolutional
autoencoder, we investgate generative adversarial network applied on autoencoders to
further improve denoising and enable more realisticmodel training condition. This includes
training with unpaired simulation and realdata, as well as training with only real detector data
without the need of simulatio. Our approach is not limited to high-purity germanium
detectors; it is broadly applicable to other detector technologies in the particle astrophysics
community and beyond.