IDEAS home Printed from https://ideas.repec.org/a/gam/jmathe/v11y2023i4p856-d1060949.html
   My bibliography  Save this article

Remote Sensing Imagery Object Detection Model Compression via Tucker Decomposition

Author

Listed:
  • Lang Huyan

    (School of Computer Science, National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, Shaanxi Provincial Key Laboratory of Speech & Image Information Processing, Northwestern Polytechnical University, Xi’an 710129, China
    Key Laboratory of Science and Technology on Space Microwave, CAST Xi’an, Xi’an 710100, China)

  • Ying Li

    (School of Computer Science, National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, Shaanxi Provincial Key Laboratory of Speech & Image Information Processing, Northwestern Polytechnical University, Xi’an 710129, China)

  • Dongmei Jiang

    (School of Computer Science, National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, Shaanxi Provincial Key Laboratory of Speech & Image Information Processing, Northwestern Polytechnical University, Xi’an 710129, China)

  • Yanning Zhang

    (School of Computer Science, National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, Shaanxi Provincial Key Laboratory of Speech & Image Information Processing, Northwestern Polytechnical University, Xi’an 710129, China)

  • Quan Zhou

    (School of Computer Science, National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology, Shaanxi Provincial Key Laboratory of Speech & Image Information Processing, Northwestern Polytechnical University, Xi’an 710129, China)

  • Bo Li

    (Key Laboratory of Science and Technology on Space Microwave, CAST Xi’an, Xi’an 710100, China)

  • Jiayuan Wei

    (Key Laboratory of Science and Technology on Space Microwave, CAST Xi’an, Xi’an 710100, China)

  • Juanni Liu

    (Key Laboratory of Science and Technology on Space Microwave, CAST Xi’an, Xi’an 710100, China)

  • Yi Zhang

    (Key Laboratory of Science and Technology on Space Microwave, CAST Xi’an, Xi’an 710100, China)

  • Peng Wang

    (Key Laboratory of Science and Technology on Space Microwave, CAST Xi’an, Xi’an 710100, China)

  • Hai Fang

    (Key Laboratory of Science and Technology on Space Microwave, CAST Xi’an, Xi’an 710100, China)

Abstract

Although convolutional neural networks (CNNs) have made significant progress, their deployment onboard is still challenging because of their complexity and high processing cost. Tensors provide a natural and compact representation of CNN weights via suitable low-rank approximations. A novel decomposed module called DecomResnet based on Tucker decomposition was proposed to deploy a CNN object detection model on a satellite. We proposed a remote sensing image object detection model compression framework based on low-rank decomposition which consisted of four steps, namely (1) model initialization, (2) initial training, (3) decomposition of the trained model and reconstruction of the decomposed model, and (4) fine-tuning. To validate the performance of the decomposed model in our real mission, we constructed a dataset containing only two classes of objects based on the DOTA and HRSC2016. The proposed method was comprehensively evaluated on the NWPU VHR-10 dataset and the CAST-RS2 dataset created in this work. The experimental results demonstrated that the proposed method, which was based on Resnet-50, could achieve up to 4.44 times the compression ratio and 5.71 times the speedup ratio with merely a 1.9% decrease in the mAP (mean average precision) of the CAST-RS2 dataset and a 5.3% decrease the mAP of the NWPU VHR-10 dataset.

Suggested Citation

  • Lang Huyan & Ying Li & Dongmei Jiang & Yanning Zhang & Quan Zhou & Bo Li & Jiayuan Wei & Juanni Liu & Yi Zhang & Peng Wang & Hai Fang, 2023. "Remote Sensing Imagery Object Detection Model Compression via Tucker Decomposition," Mathematics, MDPI, vol. 11(4), pages 1-26, February.
  • Handle: RePEc:gam:jmathe:v:11:y:2023:i:4:p:856-:d:1060949
    as

    Download full text from publisher

    File URL: https://www.mdpi.com/2227-7390/11/4/856/pdf
    Download Restriction: no

    File URL: https://www.mdpi.com/2227-7390/11/4/856/
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Ledyard Tucker, 1966. "Some mathematical notes on three-mode factor analysis," Psychometrika, Springer;The Psychometric Society, vol. 31(3), pages 279-311, September.
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Mariela González-Narváez & María José Fernández-Gómez & Susana Mendes & José-Luis Molina & Omar Ruiz-Barzola & Purificación Galindo-Villardón, 2021. "Study of Temporal Variations in Species–Environment Association through an Innovative Multivariate Method: MixSTATICO," Sustainability, MDPI, vol. 13(11), pages 1-25, May.
    2. Meyners, Michael & Qannari, El Mostafa, 2001. "Relating principal component analysis on merged data sets to a regression approach," Technical Reports 2001,47, Technische Universität Dortmund, Sonderforschungsbereich 475: Komplexitätsreduktion in multivariaten Datenstrukturen.
    3. Yuefeng Han & Rong Chen & Dan Yang & Cun-Hui Zhang, 2020. "Tensor Factor Model Estimation by Iterative Projection," Papers 2006.02611, arXiv.org, revised May 2022.
    4. DELL'ANNO, Roberto & VILLA, Stefania, 2012. "Growth in Transition Countries: Big Bang versus Gradualism," CELPE Discussion Papers 122, CELPE - CEnter for Labor and Political Economics, University of Salerno, Italy.
    5. Henk Kiers, 1991. "Hierarchical relations among three-way methods," Psychometrika, Springer;The Psychometric Society, vol. 56(3), pages 449-470, September.
    6. Willem Kloot & Pieter Kroonenberg, 1985. "External analysis with three-mode principal component models," Psychometrika, Springer;The Psychometric Society, vol. 50(4), pages 479-494, December.
    7. Pieter M. Kroonenberg & Cornelis J. Lammers & Ineke Stoop, 1985. "Three-Mode Principal Component Analysis of Multivariate Longitudinal Organizational Data," Sociological Methods & Research, , vol. 14(2), pages 99-136, November.
    8. Elisa Frutos-Bernal & Ángel Martín del Rey & Irene Mariñas-Collado & María Teresa Santos-Martín, 2022. "An Analysis of Travel Patterns in Barcelona Metro Using Tucker3 Decomposition," Mathematics, MDPI, vol. 10(7), pages 1-17, March.
    9. Xinhai Liu & Wolfgang Glänzel & Bart De Moor, 2011. "Hybrid clustering of multi-view data via Tucker-2 model and its application," Scientometrics, Springer;Akadémiai Kiadó, vol. 88(3), pages 819-839, September.
    10. Yoshio Takane & Forrest Young & Jan Leeuw, 1977. "Nonmetric individual differences multidimensional scaling: An alternating least squares method with optimal scaling features," Psychometrika, Springer;The Psychometric Society, vol. 42(1), pages 7-67, March.
    11. Dawn Iacobucci & Doug Grisaffe, 2018. "Perceptual maps via enhanced correspondence analysis: representing confidence regions to clarify brand positions," Journal of Marketing Analytics, Palgrave Macmillan, vol. 6(3), pages 72-83, September.
    12. D'Urso, Pierpaolo & Giordani, Paolo, 2003. "A least squares approach to Principal Component Analysis for interval valued data," Economics & Statistics Discussion Papers esdp03013, University of Molise, Department of Economics.
    13. Serrano Cinca, C. & Mar Molinero, C. & Gallizo Larraz, J.L., 2005. "Country and size effects in financial ratios: A European perspective," Global Finance Journal, Elsevier, vol. 16(1), pages 26-47, August.
    14. Giuseppe Brandi & Ruggero Gramatica & Tiziana Di Matteo, 2019. "Unveil stock correlation via a new tensor-based decomposition method," Papers 1911.06126, arXiv.org, revised Apr 2020.
    15. Rizzi, Alfredo & Vichi, Maurizio, 1995. "Representation, synthesis, variability and data preprocessing of a three-way data set," Computational Statistics & Data Analysis, Elsevier, vol. 19(2), pages 203-222, February.
    16. Eric Jondeau & Emmanuel Jurczenko & Michael Rockinger, 2018. "Moment Component Analysis: An Illustration With International Stock Markets," Journal of Business & Economic Statistics, Taylor & Francis Journals, vol. 36(4), pages 576-598, October.
    17. S. Lipovetsky, 2009. "Global Priority Estimation in Multiperson Decision Making," Journal of Optimization Theory and Applications, Springer, vol. 140(1), pages 77-91, January.
    18. Wilderjans, Tom & Ceulemans, Eva & Van Mechelen, Iven, 2009. "Simultaneous analysis of coupled data blocks differing in size: A comparison of two weighting schemes," Computational Statistics & Data Analysis, Elsevier, vol. 53(4), pages 1086-1098, February.
    19. Modroño Herrán, Juan Ignacio & Fernández Aguirre, María Carmen & Landaluce Calvo, M. Isabel, 2003. "Una propuesta para el análisis de tablas múltiples," BILTOKI 1134-8984, Universidad del País Vasco - Departamento de Economía Aplicada III (Econometría y Estadística).
    20. John Lastovicka, 1981. "The extension of component analysis to four-mode matrices," Psychometrika, Springer;The Psychometric Society, vol. 46(1), pages 47-57, March.

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:gam:jmathe:v:11:y:2023:i:4:p:856-:d:1060949. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: MDPI Indexing Manager (email available below). General contact details of provider: https://www.mdpi.com .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.