<?xml version='1.0' encoding='UTF-8'?>
<OAI-PMH xmlns="http://www.openarchives.org/OAI/2.0/" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/ http://www.openarchives.org/OAI/2.0/OAI-PMH.xsd">
  <responseDate>2026-06-13T15:09:24Z</responseDate>
  <request metadataPrefix="jpcoar_1.0" verb="GetRecord" identifier="oai:ipsj.ixsq.nii.ac.jp:00234182">https://ipsj.ixsq.nii.ac.jp/oai</request>
  <GetRecord>
    <record>
      <header>
        <identifier>oai:ipsj.ixsq.nii.ac.jp:00234182</identifier>
        <datestamp>2025-01-19T09:52:35Z</datestamp>
        <setSpec>1164:4619:11539:11642</setSpec>
      </header>
      <metadata>
        <jpcoar:jpcoar xmlns:datacite="https://schema.datacite.org/meta/kernel-4/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:dcndl="http://ndl.go.jp/dcndl/terms/" xmlns:dcterms="http://purl.org/dc/terms/" xmlns:jpcoar="https://github.com/JPCOAR/schema/blob/master/1.0/" xmlns:oaire="http://namespace.openaire.eu/schema/oaire/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rioxxterms="http://www.rioxx.net/schema/v2.0/rioxxterms/" xmlns:xs="http://www.w3.org/2001/XMLSchema" xmlns="https://github.com/JPCOAR/schema/blob/master/1.0/" xsi:schemaLocation="https://github.com/JPCOAR/schema/blob/master/1.0/jpcoar_scm.xsd">
          <dc:title>固有空間内におけるデータ拡張に基づく自然な不良品画像生成法の提案</dc:title>
          <dc:title xml:lang="en">A Proposal of Generating Natural Defect Images based on Data Augmentation in Eigen Space</dc:title>
          <jpcoar:creator>
            <jpcoar:creatorName>村上, 尚生</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName>平松, 直人</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName>小林, 大起</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName>秋月, 秀一</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName>橋本, 学</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName xml:lang="en">Naoki, Murakami</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName xml:lang="en">Naoto, Hiramatsu</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName xml:lang="en">Hiroki, Kobayashi</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName xml:lang="en">Shuichi, Akizuki</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:creator>
            <jpcoar:creatorName xml:lang="en">Manabu, Hashimoto</jpcoar:creatorName>
          </jpcoar:creator>
          <jpcoar:subject subjectScheme="Other">卒論スポットライトセッション (CVIM)</jpcoar:subject>
          <datacite:description descriptionType="Other">外観検査工程では，不良品サンプルが少ないという制約下での高信頼な機械学習の実現が課題である．これに対し，良品画像と不良品画像を組み合わせるデータ拡張が一般的であるが，生成される欠陥と対象物の境界が不自然な不良品画像が生成されるという問題がある．そこで，本研究では，データ拡張を主成分分析によって生成した固有空間内でおこなうことによって，欠陥と対象物の境界が自然な不良品画像を生成する手法を提案する．また，提案手法により生成された不良品画像を識別モデルの学習に用いた場合，薬品カプセルデータセットでのクラス識別正解率が 96.3% であり，従来手法を 20.6 ポイント上回っていることを確認した．</datacite:description>
          <datacite:description descriptionType="Other">In visual inspection, it is challenging to achieve high-reliability machine learning when there are only a few defective samples. In previous data augmentation, it is common to combine normal images and defect images, but it has the problem of generating defect images with unnatural boundaries between the defects and the object. In this study, we propose a method to generate defect images with natural boundaries between defects and objects by performing data augmentation in the Eigen space generated by principal component analysis. When using defect images generated by the proposed method for training the discriminator, the accuracy of normal/defect discrimination achieved 96.3% for the capsule dataset. This result exceeded the previous methods by 20.6%.</datacite:description>
          <dc:publisher xml:lang="ja">情報処理学会</dc:publisher>
          <datacite:date dateType="Issued">2024-05-08</datacite:date>
          <dc:language>jpn</dc:language>
          <dc:type rdf:resource="http://purl.org/coar/resource_type/c_18gh">technical report</dc:type>
          <jpcoar:identifier identifierType="URI">https://ipsj.ixsq.nii.ac.jp/records/234182</jpcoar:identifier>
          <jpcoar:sourceIdentifier identifierType="ISSN">2188-8701</jpcoar:sourceIdentifier>
          <jpcoar:sourceIdentifier identifierType="NCID">AA11131797</jpcoar:sourceIdentifier>
          <jpcoar:sourceTitle>研究報告コンピュータビジョンとイメージメディア（CVIM）</jpcoar:sourceTitle>
          <jpcoar:volume>2024-CVIM-238</jpcoar:volume>
          <jpcoar:issue>51</jpcoar:issue>
          <jpcoar:pageStart>1</jpcoar:pageStart>
          <jpcoar:pageEnd>6</jpcoar:pageEnd>
          <jpcoar:file>
            <jpcoar:URI label="IPSJ-CVIM24238051.pdf">https://ipsj.ixsq.nii.ac.jp/record/234182/files/IPSJ-CVIM24238051.pdf</jpcoar:URI>
            <jpcoar:mimeType>application/pdf</jpcoar:mimeType>
            <jpcoar:extent>1.2 MB</jpcoar:extent>
            <datacite:date dateType="Available">2026-05-08</datacite:date>
          </jpcoar:file>
        </jpcoar:jpcoar>
      </metadata>
    </record>
  </GetRecord>
</OAI-PMH>
