Wire-Cell Pattern Recognition for Liquid Argon Time Projection Chambers

Liquid argon time projection chambers (LArTPCs) are widely used in current and future neutrino experiments to answer some of the key questions about neutrino physics. Event reconstruction is a critical but challenging task in analyzing the data from LArTPCs. Following the principle of LArTPC, we developed a new tomographic event reconstruction paradigm, Wire-Cell. Wire-Cell achieved fully automated LArTPC reconstruction with multiple components including noise filtering, signal processing, 3D charge reconstruction and clustering, light-charge matching and higher-level pattern recognition. In this talk, we present the Wire-Cell pattern recognition algorithms incorporating both traditional and deep learning techniques.