Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE
We present the performance of a semantic segmentation network, SparseSSNet, that provides pixel-level classification of MicroBooNE data. The MicroBooNE experiment employs a liquid argon time projection chamber for the study of neutrino properties and interactions. SparseSSNet is a submanifold spa...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
American Physical Society (APS)
2022
|
| Online Access: | https://hdl.handle.net/1721.1/141997 |
| _version_ | 1826190261555822592 |
|---|---|
| author | Conrad, Janet Hen, Or |
| author2 | Massachusetts Institute of Technology. Department of Physics |
| author_facet | Massachusetts Institute of Technology. Department of Physics Conrad, Janet Hen, Or |
| author_sort | Conrad, Janet |
| collection | MIT |
| description | We present the performance of a semantic segmentation network, SparseSSNet,
that provides pixel-level classification of MicroBooNE data. The MicroBooNE
experiment employs a liquid argon time projection chamber for the study of
neutrino properties and interactions. SparseSSNet is a submanifold sparse
convolutional neural network, which provides the initial machine learning based
algorithm utilized in one of MicroBooNE's $\nu_e$-appearance oscillation
analyses. The network is trained to categorize pixels into five classes, which
are re-classified into two classes more relevant to the current analysis. The
output of SparseSSNet is a key input in further analysis steps. This technique,
used for the first time in liquid argon time projection chambers data and is an
improvement compared to a previously used convolutional neural network, both in
accuracy and computing resource utilization. The accuracy achieved on the test
sample is $\geq 99\%$. For full neutrino interaction simulations, the time for
processing one image is $\approx$ 0.5 sec, the memory usage is at 1 GB level,
which allows utilization of most typical CPU worker machine. |
| first_indexed | 2024-09-23T08:37:33Z |
| format | Article |
| id | mit-1721.1/141997 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T08:37:33Z |
| publishDate | 2022 |
| publisher | American Physical Society (APS) |
| record_format | dspace |
| spelling | mit-1721.1/1419972023-02-08T20:55:38Z Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE Conrad, Janet Hen, Or Massachusetts Institute of Technology. Department of Physics We present the performance of a semantic segmentation network, SparseSSNet, that provides pixel-level classification of MicroBooNE data. The MicroBooNE experiment employs a liquid argon time projection chamber for the study of neutrino properties and interactions. SparseSSNet is a submanifold sparse convolutional neural network, which provides the initial machine learning based algorithm utilized in one of MicroBooNE's $\nu_e$-appearance oscillation analyses. The network is trained to categorize pixels into five classes, which are re-classified into two classes more relevant to the current analysis. The output of SparseSSNet is a key input in further analysis steps. This technique, used for the first time in liquid argon time projection chambers data and is an improvement compared to a previously used convolutional neural network, both in accuracy and computing resource utilization. The accuracy achieved on the test sample is $\geq 99\%$. For full neutrino interaction simulations, the time for processing one image is $\approx$ 0.5 sec, the memory usage is at 1 GB level, which allows utilization of most typical CPU worker machine. 2022-04-21T12:43:25Z 2022-04-21T12:43:25Z 2021 2022-04-21T12:39:43Z Article http://purl.org/eprint/type/JournalArticle https://hdl.handle.net/1721.1/141997 Conrad, Janet and Hen, Or. 2021. "Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE." Physical Review D, 103 (5). en 10.1103/PHYSREVD.103.052012 Physical Review D Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf American Physical Society (APS) APS |
| spellingShingle | Conrad, Janet Hen, Or Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE |
| title | Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE |
| title_full | Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE |
| title_fullStr | Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE |
| title_full_unstemmed | Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE |
| title_short | Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE |
| title_sort | semantic segmentation with a sparse convolutional neural network for event reconstruction in microboone |
| url | https://hdl.handle.net/1721.1/141997 |
| work_keys_str_mv | AT conradjanet semanticsegmentationwithasparseconvolutionalneuralnetworkforeventreconstructioninmicroboone AT henor semanticsegmentationwithasparseconvolutionalneuralnetworkforeventreconstructioninmicroboone |