Hit-level separation of track-like particles within dense electromagnetic showers

Develop and validate improved hit-level algorithms for separating track-like particles traveling through dense electromagnetic showers in liquid argon time projection chamber (LArTPC) detectors, achieving higher performance than existing methods to support downstream reconstruction tasks.

Background

Separating track-like and shower-like topologies at the hit level is a critical step for reconstruction frameworks such as Pandora in LArTPC experiments, enabling particle identification and event reconstruction. Multiple machine learning approaches have been explored.

Despite these efforts, the authors explicitly state that separating track-like particles within dense electromagnetic showers remains an open reconstruction problem where improved performance is needed.