ProtoDUNE-SP is a single-phase liquid argon time projection chamber (LArTPC) built at the CERN neutrino platform. It serves as a prototype for the upcoming Deep Underground Neutrino Experiment (DUNE) to study charged-particle interactions under a test beam environment. In this poster, we outline a machine learning (ML) based event reconstruction for building a 3-dimensional image from the projections of wire plane readouts, classifying particle types at the pixel level, locating start/end points of particles, and clustering of electromagnetic showers. We will present a study of applying the above ML techniques to reconstruct neutral pion candidates in ProtoDUNE-SP.
A novel approach to reconstruct neutral pion in ProtoDUNE-SP detector