Adrien Hourlier (MIT) : 3D track finding for MicroBooNE’s deep learning based event reconstruction chain

MicroBooNE is a Liquid Argon Time Projection Chamber (LArTPC) 
neutrino experiment on the Booster Neutrino Beamline at the Fermi 
National Accelerator Laboratory, with an 85-tonne active mass. One 
of MicroBooNE's primary physics goals is to investigate the excess 
of electron neutrino events seen by MiniBooNE in the [200-600] MeV 
range. MicroBooNE will constrain the intrinsic electron neutrino 
component of the beam by measuring the muon neutrino spectrum. Our 
low-energy excess analysis makes use of deep learning algorithms 
applied to the high-resolution images provided by the MicroBooNE 
LArTPC. I will present a novel 3D event reconstruction based on 
computer vision tools and a stochastic search algorithm, yielding 
an excellent energy resolution for 1mu1p muon neutrino 
interactions in the [200-1500] MeV range. I will then present 
validation studies verifying the good agreement of our simulation 
to our muon neutrino data.
Your browser is out of date!

Update your browser to view this website correctly. Update my browser now