Johannes Bilk ; Stephanie Käs ; Sören Lange ; Elisabetta Prencipe ; Timo Schellhaas

26 February 2024
DPG-Frühjahrstagung Karlsruhe

Abstract: The identification of slow pions in Belle II experiments presents a no- table challenge, arising from their high dE/dx energy loss and their short flight path in the tracking detectors. This study introduces a method employing advanced machine learning algorithms to accurately detect pions with momentum p<100 MeV/c exclusively with the Belle II pixeldetector (PXD). By analyzing detector signals (in particular a 9x9 pixel matrix) with image processing and pattern recognition methods, this approach significantly boosts the efficiency and accu- racy. Offline and online (FPGA) implemenation will be discussed.