English  |  正體中文  |  简体中文  |  Post-Print筆數 : 27 |  Items with full text/Total items : 110180/141115 (78%)
Visitors : 46604905      Online Users : 862
RC Version 6.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
    •  Adv. Search
    Home ‧ Login ‧ Upload ‧ Help ‧ About ‧ Administer Goto mobile version
    政大機構典藏 > 理學院 > 應用數學系 > 會議論文 >  Item 140.119/139050


    Please use this identifier to cite or link to this item: https://nccur.lib.nccu.edu.tw/handle/140.119/139050


    Title: Improve the LSTM and GRU model for small training data by wavelet transformation
    Authors: 曾正男
    Tzeng, Jengnan
    Lai, Yen-Ru
    Lin, Ming-Lai
    Lin, Yu-Han
    Shih, Yu-Cheng
    Contributors: 應數系
    Keywords: Wavelet transforms;Logic gates;Artificial intelligence;Training;Training data;Cameras;alarm systems;artificial intelligence;automobiles;CCD image sensors;collision avoidance;image representation;image resolution;motorcycles;radar detection;recurrent neural nets;road safety;traffic engineering computing;wavelet transforms;infrared rays;car reversing radar;GRU model;LSTM;AI technology;image representation;image AI;low power consumption requirements;high-precision prediction;anti-collision warnings;low-cost anti-collision technology;artificial intelligence technology;low-resolution CCD;high-temperature environments;360-degree feedback;radar systems;radar detection;collision prediction technology;wavelet transformation;small training data;development costs;wavelet;Haar basis;AI;ADAS;small training data
    Date: 2020-07
    Issue Date: 2022-02-10 15:00:25 (UTC+8)
    Abstract: Regarding collision prediction technology, the most common are car reversing radar and infrared rays, which provide warnings by sensing the distance between objects and cars. Although radar detection is accurate, radar systems that can provide instant 360-degree feedback are very expensive. The cost of infrared is much lower, but they cannot be applied to high-temperature environments. As the result, the technology of preventing collisions using only images from cameras is an important artificial intelligence topic in recent years. If a low-resolution CCD and artificial intelligence technology can be used to achieve a certain degree of accuracy in collision prediction, then low-cost anti-collision technology is worth looking forward to. Furthermore, we hope to provide anti-collision warnings on motorcycles and bicycles using this technology. In order to achieve this goal, computational simplification is a technical threshold. It`s only when simple calculations achieve high-precision prediction can we meet the low power consumption requirements for image AI to be applied small vehicles. Therefore, we hope to find out a better image representation basis and combine it with AI technology to fulfill the requirements of less calculation and high accuracy. In addition, we also hope to create models with sufficient accuracy with small training data. This experiment will reduce development costs and get better efficiency in the early stage of developing ADAS.
    Relation: 2020 International Joint Conference on Neural Networks (IJCNN), IEEE
    Data Type: conference
    DOI 連結: https://doi.org/10.1109/IJCNN48605.2020.9206840
    DOI: 10.1109/IJCNN48605.2020.9206840
    Appears in Collections:[應用數學系] 會議論文

    Files in This Item:

    File Description SizeFormat
    340.pdf866KbAdobe PDF2169View/Open


    All items in 政大典藏 are protected by copyright, with all rights reserved.


    社群 sharing

    著作權政策宣告 Copyright Announcement
    1.本網站之數位內容為國立政治大學所收錄之機構典藏，無償提供學術研究與公眾教育等公益性使用，惟仍請適度，合理使用本網站之內容，以尊重著作權人之權益。商業上之利用，則請先取得著作權人之授權。
    The digital content of this website is part of National Chengchi University Institutional Repository. It provides free access to academic research and public education for non-commercial use. Please utilize it in a proper and reasonable manner and respect the rights of copyright owners. For commercial use, please obtain authorization from the copyright owner in advance.

    2.本網站之製作，已盡力防止侵害著作權人之權益，如仍發現本網站之數位內容有侵害著作權人權益情事者，請權利人通知本網站維護人員(nccur@nccu.edu.tw)，維護人員將立即採取移除該數位著作等補救措施。
    NCCU Institutional Repository is made to protect the interests of copyright owners. If you believe that any material on the website infringes copyright, please contact our staff(nccur@nccu.edu.tw). We will remove the work from the repository and investigate your claim.
    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - Feedback