skip to main content
10.1145/3674225.3674261acmotherconferencesArticle/Chapter ViewAbstractPublication PagespeaiConference Proceedingsconference-collections
research-article

A Lightweight Transmission Line Defect Detection Method via Joint Optimization of Pruning and Quantization

Published: 31 July 2024 Publication History

Abstract

The normal operation of power transmission lines is critical for ensuring electricity supply and economic growth. However, various factors like weather, human activity, wildlife, and wildfires can disrupt operations. Artificial intelligence methods for visual defect analysis offer higher accuracy than traditional techniques but are computationally expensive. This study proposes an efficient approach for on-device analysis by jointly optimizing quantization and pruning of models to reduce errors and accelerate inference, enabling fast diagnosis of hazards in transmission line infrastructure. The method achieves 3.3 times model acceleration to support front-end deployment, aiding timely identification of risks that threaten stable delivery of electricity.

References

[1]
Wu, Jun, "Detection method based on improved faster R-CNN for pin defect in transmission lines." E3S Web of Conferences. Vol. 300. EDP Sciences, 2021.
[2]
Wu, Tao, "Insulator defect detection method for lightweight YOLOV3." Computer Engineering 45.8. 2019, 275-280.
[3]
LECUN Y, DENKER J S, SOLLA S A. Optimal brain damage[C]//TOURETZKY D S. Advances in Neural Information Processing Systems 2. Denver, Colorado, USA: Morgan Kaufmann, 1989, 598-605.
[4]
HASSIBI B, STORK D G. Second order derivatives for network pruning: Optimal brain surgeon [C]//HANSON S J, COWAN J D, GILES C L. Advances in Neural Information Processing Systems 5. Denver, Colorado, USA: Morgan Kaufmann, 1992, 164-171.
[5]
HAN S, POOL J, TRAN J, Learning both weights and connections for efficient neural network [C]//CORTES C, LAWRENCE N D, LEE D D, Advances in Neural Information Processing Systems 28. Montreal, Quebec, Canada, 2015, 1135-1143.
[6]
Yiwen Guo, Anbang Yao, and Yurong Chen. 2016. Dynamic network surgery for efficient DNNs. In Proceedings of the 30th International Conference on Neural Information Processing Systems (NIPS'16). Curran Associates Inc., Red Hook, NY, USA, 1387–1395.
[7]
LEBEDEV V, LEMPITSKY V S. Fast convnets using group-wise brain damage[C/OL]// 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, USA: IEEE Computer Society, 2016, 2554-2564.
[8]
HU H, PENG R, TAI Y W, Network trimming: A data-driven neuron pruning approach towards efficient deep architectures [J]. arXiv preprint arXiv:1607.03250, 2016.
[9]
LUO J H, WU J, LIN W. Thinet: A filter level pruning method for deep neural network compression [J]. arXiv preprint arXiv:1707.06342, 2017.
[10]
HE Y, ZHANG X, SUN J. Channel pruning for accelerating very deep neural networks[J]. arXiv preprint arXiv:1707.06168, 2017.
[11]
LIU Z, LI J, SHEN Z, Learning efficient convolutional networks through network slimming[J]. arxiv preprint, 2017, 1708.
[12]
GUPTA S, AGRAWAL A, GOPALAKRISHNAN K, Deep learning with limited numerical precision [C]//BACH F, BLEI D. Proceedings of Machine Learning Research: volume 37 Proceedings of the 32nd International Conference on Machine Learning. Lille, France: PMLR, 2015, 1737-1746.
[13]
QIU J, WANG J, YAO S, Going deeper with embedded fpga platform for convolutional neural network [C/OL]//CHEN D, GREENE J W. FPGA ’16: Proceedings of the 2016 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays. New York, NY, USA: ACM, 2016, 26-35.
[14]
COURBARIAUX M, BENGIO Y, DAVID J. Binaryconnect: Training deep neural networks with binary weights during propagations[C]//CORTES C, LAWRENCE N D, LEE D D, Advances in Neural Information Processing Systems 28. Montreal, Quebec, Canada, 2015, 3123-3131.
[15]
LIN Z, COURBARIAUX M, MEMISEVIC R, Neural networks with few multiplications[J/OL]. CoRR, 2015, abs/1510.03009. http://arxiv.org/abs/1510.03009.
[16]
RASTEGARI M, ORDONEZ V, REDMON J, Xnornet: Imagenet classification using binary convolutional neural networks[C]//LEIBE B, MATAS J, SEBE N, Computer Vision - ECCV 2016. Cham: Springer International Publishing, 2016, 525-542.
[17]
CAI Z, HEX, SUN J, Deep learning with low precision by half-wave gaussian quantization [C/OL]//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, HI, USA: IEEE Computer Society, 2017, 5406-5414.
[18]
DONG Y, NI R, LI J, Learning accurate low-bit deep neural networks with stochastic quantization [J/OL]. CoRR, 2017, abs/1708.01001. http://arxiv.org/abs/1708.01001.

Index Terms

  1. A Lightweight Transmission Line Defect Detection Method via Joint Optimization of Pruning and Quantization

    Recommendations

    Comments

    Information & Contributors

    Information

    Published In

    cover image ACM Other conferences
    PEAI '24: Proceedings of the 2024 International Conference on Power Electronics and Artificial Intelligence
    January 2024
    969 pages
    ISBN:9798400716638
    DOI:10.1145/3674225
    Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than the author(s) must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected].

    Publisher

    Association for Computing Machinery

    New York, NY, United States

    Publication History

    Published: 31 July 2024

    Permissions

    Request permissions for this article.

    Check for updates

    Qualifiers

    • Research-article
    • Research
    • Refereed limited

    Conference

    PEAI 2024

    Contributors

    Other Metrics

    Bibliometrics & Citations

    Bibliometrics

    Article Metrics

    • 0
      Total Citations
    • 10
      Total Downloads
    • Downloads (Last 12 months)10
    • Downloads (Last 6 weeks)1
    Reflects downloads up to 30 Jan 2025

    Other Metrics

    Citations

    View Options

    Login options

    View options

    PDF

    View or Download as a PDF file.

    PDF

    eReader

    View online with eReader.

    eReader

    HTML Format

    View this article in HTML Format.

    HTML Format

    Figures

    Tables

    Media

    Share

    Share

    Share this Publication link

    Share on social media