Jul 16 – 26, 2022
US/Pacific timezone

A novel ML-based method of primary vertex reconstruction in high pile-up condition

Jul 18, 2022, 7:00 PM
2h 20m
211 South Ballroom (HUB)

211 South Ballroom



Haoran Zhao (University of Washington)


The High-Luminosity LHC (HL-LHC) is expected to reach the peak instantaneous luminosity of $7.5×10^{34}\mathrm{cm}^{−2}\mathrm{s}^{−1}$ at a center-of-mass energy of $\sqrt{s}$= 14 TeV. This leads to an extremely high density environment with up to 200 interactions per proton-proton bunch crossing. Under these conditions event reconstruction represents a major challenge for experiments due to the high pileup vertices present.
To tackle the dense environment, we adapted a novel ML-based method named Sparse Point-Voxel Convolution Neural Network(SPVCNN), the current state-of-the-art techniques in computer vision, which leverages point-based method and space voxelization to categorize tracks into primary vertices.
In this poster, the performance of SPVCNN vertexing will be presented, as well as the comparison with the conventional Adaptive Multi-Vertex Finding(AMVF) algorithm.

In-person or Virtual? In-person

Primary authors

Haoran Zhao (University of Washington) Alexander Schuy (University of Washington) Shih-Chieh Hsu (University of Washington) Philip Harris (MIT) Mr Ke Li (University of Washington) Mr Zhijian Liu (MIT) Dr Song Han (MIT)

Presentation materials