Deep Convolutional Neural Networks (CNNs) have exhibited remarkable efficacy in data analysis across various domains of physical research. Within the realm of particle and high-energy physics, an imperative challenge revolves around the scrutiny of particle tracks acquired through track detectors, exemplified by liquefied argon time projection chambers (LArTPCs). These cutting-edge neural networks have showcased unparalleled performance in tackling this intricate task, demonstrating their profound potential in enhancing particle track analysis.
Since the scalability requirement is imposed on neutrino detectors, it is also imposed on CNN model implementations. The scalability for LArTPC data processing is achieved due to the sparsity of the data, which have the form of thin trajectories. To process sparse data, subspecies sparse convolutional networks (SSCNs) and sparse tensor networks (STNs) have been proposed.
In this paper, we present the results of semantic segmentation of the LArTPC-simulated PILArNET data using STNs and test various modifications of the classical U-Net architecture: Attention U-ResNet, U-ResNet3+, U-ResNet with an additional deep supervision block, as well as loss functions: focal loss, balanced cross-entropy to improve the accuracy of Michel electron identification. The best results were obtained for balanced cross-entropy. Further improvement is possible by combining several methods.
Key words: neutrino, classification, neural network, machine learning.
Full text
References
- P. Benetti et al., Nucl. Instrum. Methods Phys. Res. A 332, 395 (1993);
Crossref
- R. Acciarri et al. (MicroBooNE Collaboration), J. Instrum. 12, P02017 (2017);
Crossref
- M. Antonello et al., preprint arXiv:1503.01520 (2015);
Crossref
- S. Amerio et al. (ICARUS Collaboration), Nucl. Instrum. Methods Phys. Res. A 527, 329 (2004);
Crossref
- R. Acciarri et al., preprint arXiv:1601.02984 (2016);
Crossref
- F. Drielsma, K. Terao, L. Dominé, D. H. Koh, preprint arXiv.2102.01033 (2021);
Crossref
- C. Choy, JY. Gwak, S. Savarese, in 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (Long Beach, CA, USA, 2019), p. 3070;
Crossref
- B. Graham, L. van der Maaten, preprint arXiv:1706.01307 (2017);
Crossref
- B. Graham, M. Engelcke, L. van der Maaten, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (IEEE, 2018), p. 9224;
http://openaccess.thecvf.com/content_cvpr_2018/html/Graham_3D_Semantic_Segmentation_CVPR_2018_paper.html
- L. Dominé, K. Terao (DeepLearnPhysics Collaboration), Phys. Rev. D 102, 012005 (2020);
Crossref
- C. Adams, K. Terao, T. Wongjirad, preprint arXiv: 2006.01993 (2020);
Crossref
- D. H. Koh et al., preprint arXiv:2007.03083 (2020);
Crossref
- F. Drielsma et al., preprint arXiv:2007.01335 (2020);
Crossref
- L. Dominé et al., Phys. Rev. D 104, 032004 (2021);
Crossref
- S. Agostinelli et al. (GEANT4), Nucl. Instrum. Meth. A 506, 250 (2003);
Crossref
- E. L. Snider, G. Petrillo, J. Phys. Conf. Ser. 898, 042057 (2017);
Crossref
- S. J. Russell, P. Norvig, Artificial Intelligence: A Modern Approach, 3rd ed. (Prentice Hall, 2010).
- O. Ronneberger, P. Fischer, T. Brox, preprint arXiv:1505.04597 (2015);
Crossref
- M. J. Berger, J. S. Coursey, M. A. Zucker, J. Chang, ESTAR, PSTAR, and ASTAR: Computer Programs for Calculating Stopping-Power and Range Tables for Electrons, Protons, and Helium Ions (version 1.2.3) (National Institute of Standards and Technology, 2005);
Crossref
- R. Aciarri et al., J. Instrum. 12, P09014 (2017);
Crossref
- Adam Abed Abud on behalf of the DUNE collab., J. Phys. Conf. Ser. 2438, 012125 (2023);
Crossref
- T.-Y. Lin, P. Goyal, R. Girshick, K. M. He, P. Dollar, preprint arXiv:1708.02002 (2017);
Crossref
- H. Huang et al., preprint arXiv:1708.02002 (2020);
Crossref
- O. Oktay et al., preprint arXiv:1804.03999 (2018);
Crossref
- L. Michel, Proc. Phys. Soc. A 63 514 (1950);
Crossref
- L. Michel, Proc. Phys. Soc. A 63, 1371 (1950);
Crossref
- M. Bardon et al., Phys. Rev. Lett. 14, 449 (1965);
Crossref
- R. Acciarri et al., J. Instrum. 12, P09014 (2017);
Crossref