1. What are the contributions mentioned in the paper "Semantic segmentation with a sparse convolutional neural network for event" ?
In this paper, the authors present a survey of the state-of-the-art work in this area.
read more
Semantic Segmentation with a Sparse Convolutional Neural Network for Event Reconstruction in MicroBooNE
TL;DR: SparseSSNet as discussed by the authors is a submanifold sparse convolutional neural network, which provides the initial machine learning based algorithm utilized in one of MicroBooNE's appearance oscillation analyses.
read more
Abstract: 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.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Most frequently asked questions
Figures
FIG. 11: Five-class confusion matrices for the second induction plane test sample. Each box represents the fraction of pixels which are from the class stated on x-axis and predicted as class stated in y-axis. The smallest number of pixels is O(105) for Michel electrons. All other classes vary between 106 − 107 pixels. 
FIG. 10: Five-class confusion matrices for the first induction plane test sample. Each box represents the fraction of pixels which are from the class stated on x-axis and predicted as class stated in y-axis. The smallest number of pixels is O(105) for Michel electrons. All other classes vary between 106 − 107 pixels. 
TABLE III: SparseSSNet track and shower accuracy for the test sample and the neutrino interaction central value simulation samples (both full-BNB and intrinsic νe). The results are obtained from the two induction planes. The number of pixels associated with each class is O(107) pixels except for the full-BNB shower which is O(105). The drop in the shower accuracy for the neutrino interaction sample is explained in V B. 
FIG. 8: An example of a simulated νe interaction projected on the first induction plane. This is a (300 × 300) pixels crop from the full detector image shown in Fig. 12. (a) Pixel intensity of interaction overlayed with cosmic rays. (b) The produced particles upon interaction, before overlaying cosmic rays. (c) SparseSSNet predictions. Notice that although some shower pixels are mis-classified, the fraction of shower-like pixels is larger than 0.2 and therefore this interaction will be correctly classified as νe interaction (see Sec. V C). 
FIG. 9: An example of a simulated νµ interaction projected on the collection plane. This is a (300 × 300) pixels crop from the full detector image shown in Fig. 13. (a) Pixel intensity of interaction overlayed with cosmic rays. (b) The produced particles upon interaction, before overlaying cosmic rays. (c) SparseSSNet predictions. 
FIG. 13: An example of a simulated νµ interaction projected on the first induction plane. This is the full detector image of the νe interaction shown in Fig. 9. (a) Pixel intensity of interaction overlayed with cosmic rays. (b) The label assigned to the simulated neutrino interaction. As cosmic rays are taken from beam-off data they are not assigned with labels. (c) SparseSSNet’s predictions.
Citations
Machine learning in the search for new fundamental physics
G. Karagiorgi,Bowen Wang,Anna Bondarenko,Nabeel Taha Ali Belal , Khalid Akbar Abdullah,Sayed Khalil Kohi +4 more
TL;DR: A review of the state-of-the-art methods and applications for new physics searches in the context of terrestrial high-energy physics experiments, including the Large Hadron Collider, rare event searches and neutrino experiments, can be found in this paper .
A Convolutional Neural Network for Multiple Particle Identification in the MicroBooNE Liquid Argon Time Projection Chamber
P. Abratenko,M. Alrashed,R. An,J. Anthony,J. Asaadi,Adi Ashkenazi,S. Balasubramanian,B. Baller,C. Barnes,G.D. Barr,V. Basque,L. Bathe-Peters,O. Benevides Rodrigues,S. Berkman,A. Bhanderi,A. Bhat,M. Bishai,Andrew Blake,T. A. Bolton,L. Camilleri,D. Caratelli,I. Caro Terrazas,R. Castillo Fernandez,F. Cavanna,Giuseppe Benedetto Cerati,Yi Chen,E. Church,D. Cianci,Jan Conrad,M. E. Convery,L. Cooper-Troendle,J. I. Crespo-Anadón,M. Del Tutto,D. Devitt,R. Diurba,Laura Dominé,R. Dorrill,K. E. Duffy,S. A. Dytman,B. Eberly,Antonio Ereditato,L. Escudero Sanchez,John Evans,G. A. Fiorentini Aguirre,R. S. Fitzpatrick,B. T. Fleming,N. Foppiani,D. Franco,A. P. Furmanski,D. Garcia-Gamez,Steven Gardiner,G. Ge,S. Gollapinni,O. Goodwin,E. Gramellini,Paul J. Green,H. Greenlee,W. Q. Gu,R. Guenette,P. Guzowski,E. D. Hall,P. M. Hamilton,Or Hen,G. A. Horton-Smith,A. Hourlier,E.-C. Huang,R. Itay,C. W. James,J. Jan de Vries,Xiaolu Ji,L. Jiang,J. H. Jo,R. A. Johnson,Y.-J. Jwa,N. Kamp,G. Karagiorgi,W. Ketchum,B. Kirby,M. H. Kirby,T. Kobilarcik,I. Kreslo,R. LaZur,I. Lepetic,K. Li,Yifan Li,B. R. Littlejohn,D. Lorca,W. C. Louis,X. Luo,A. Marchionni,S. Marcocci,C. Mariani,D. Marsden,Jennifer L. Marshall,J. Martin-Albo,D. A. Martinez Caicedo,K. Mason,A. Mastbaum,N. McConkey,V. Meddage,T. Mettler,Kate C. Miller,J. C. Mills,K. P. Mistry,A. Mogan,T. A. Mohayai,J. Moon,Margaret M. Mooney,A. F. Moor,C. D. Moore,J. Mousseau,Michael T. Murphy,D. Naples,A. Navrer-Agasson,R. K. Neely,P. Nienaber,J. A. Nowak,O. Palamara,V. Paolone,Afroditi Papadopoulou,V. Papavassiliou,S. F. Pate,A. Paudel,Z. Pavlovic,E. Piasetzky,I. Ponce-Pinto,D. Porzio,Sebastien Prince,Xin Qian,J. L. Raaf,Veljko Radeka,A. Rafique,M. Reggiani-Guzzo,L. Ren,L.S. Rochester,J. Rodriguez Rondon,H. E. Rogers,M. Rosenberg,M. Ross-Lonergan,B. Russell,G. Scanavini,D.W. Schmitz,A. Schukraft,M. H. Shaevitz,R. Sharankova,J. R. Sinclair,A. M. Smith,E.L. Snider,M. Soderberg,S. Söldner-Rembold,S.R. Soleti,Panagiotis Spentzouris,J. Spitz,M. Stancari,J. St. John,Thomas Strauss,K. Sutton,S. Sword-Fehlberg,A. M. Szelc,N. Tagg,William Tang,Kazuhiro Terao,C. Thorpe,M. Toups,Y. T. Tsai,Serhan Tufanli,M. A. Uchida,T. L. Usher,W. Van De Pontseele,B. Viren,M. Weber,H. Y. Wei,Z. Williams,S. Wolbers,T. Wongjirad,M. Wospakrik,Wei Wu,T. Yang,G. Yarbrough,L.E. Yates,G. P. Zeller,J. Zennamo,C. Zhang +182 more
TL;DR: In this paper, a convolutional neural network (CNN) for multiple object classification was proposed. And the multiple particle identification (MPID) network, a CNN-based approach for multiple particle detection, is presented.
41
Heavy neutral leptons below the kaon mass at hodoscopic neutrino detectors
11 May 2022
TL;DR: In this article , the authors show that the T2K and MicroBooNE experiments provide the best limits on the mixing of neutrinos with muon-neutrinos, outperforming past-generation experiments.
Search for an Excess of Electron Neutrino Interactions in MicroBooNE Using Multiple Final-State Topologies
13 Jun 2022
TL;DR: In this paper , a measurement of ν-e interactions from the Fermilab Booster Neutrino Beam using the MicroBooNE liquid argon time projection chamber is presented.
References
Deep learning
TL;DR: Deep learning is making major advances in solving problems that have resisted the best attempts of the artificial intelligence community for many years, and will have many more successes in the near future because it requires very little engineering by hand and can easily take advantage of increases in the amount of available computation and data.
67K
Geant4—a simulation toolkit
S. Agostinelli,John Allison,K. Amako,J. Apostolakis,Henrique Araujo,P. Arce,Makoto Asai,D. Axen,S. Banerjee,G. Barrand,F. Behner,Lorenzo Bellagamba,J. Boudreau,L. Broglia,A. Brunengo,H. Burkhardt,Stephane Chauvie,J. Chuma,R. Chytracek,Gene Cooperman,G. Cosmo,P. V. Degtyarenko,Andrea Dell'Acqua,G. Depaola,D. Dietrich,R. Enami,A. Feliciello,C. Ferguson,H. Fesefeldt,Gunter Folger,Franca Foppiano,Alessandra Forti,S. Garelli,S. Gianì,R. Giannitrapani,D. Gibin,J. J. Gomez Y Cadenas,I. González,G. Gracia Abril,G. Greeniaus,Walter Greiner,Vladimir Grichine,A. Grossheim,Susanna Guatelli,P. Gumplinger,R. Hamatsu,K. Hashimoto,H. Hasui,A. Heikkinen,A. S. Howard,Vladimir Ivanchenko,A. Johnson,F.W. Jones,J. Kallenbach,Naoko Kanaya,M. Kawabata,Y. Kawabata,M. Kawaguti,S.R. Kelner,Paul R. C. Kent,A. Kimura,T. Kodama,R. P. Kokoulin,M. Kossov,Hisaya Kurashige,E. Lamanna,Tapio Lampén,V. Lara,Veronique Lefebure,F. Lei,M. Liendl,W. S. Lockman,Francesco Longo,S. Magni,M. Maire,E. Medernach,K. Minamimoto,P. Mora de Freitas,Yoshiyuki Morita,K. Murakami,M. Nagamatu,R. Nartallo,Petteri Nieminen,T. Nishimura,K. Ohtsubo,M. Okamura,S. W. O'Neale,Y. Oohata,K. Paech,J Perl,Andreas Pfeiffer,Maria Grazia Pia,F. Ranjard,A.M. Rybin,S.S Sadilov,E. Di Salvo,Giovanni Santin,Takashi Sasaki,N. Savvas,Y. Sawada,Stefan Scherer,S. Sei,V. Sirotenko,David J. Smith,N. Starkov,H. Stoecker,J. Sulkimo,M. Takahata,Satoshi Tanaka,E. Tcherniaev,E. Safai Tehrani,M. Tropeano,P. Truscott,H. Uno,L. Urbán,P. Urban,M. Verderi,A. Walkden,W. Wander,H. Weber,J.P. Wellisch,Torre Wenaus,D.C. Williams,Douglas Wright,T. Yamada,H. Yoshida,D. Zschiesche +126 more
TL;DR: The Gelfant 4 toolkit as discussed by the authors is a toolkit for simulating the passage of particles through matter, including a complete range of functionality including tracking, geometry, physics models and hits.
The class imbalance problem: A systematic study
Nathalie Japkowicz,Shaju Stephen +1 more
- 01 Oct 2002
TL;DR: The assumption that the class imbalance problem does not only affect decision tree systems but also affects other classification systems such as Neural Networks and Support Vector Machines is investigated.
3.4K
Related Papers (5)
Chih-Chung Hsu,Chia-Wen Lin +1 more
- 01 Sep 2017
Pietro Zanuttigh,Ludovico Minto +1 more
- 01 Sep 2017