<?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>2026-05-20T23:38:20Z</responseDate>
  <request verb="GetRecord" metadataPrefix="oai_dc" identifier="oai:ipsj.ixsq.nii.ac.jp:00234836">https://ipsj.ixsq.nii.ac.jp/oai</request>
  <GetRecord>
    <record>
      <header>
        <identifier>oai:ipsj.ixsq.nii.ac.jp:00234836</identifier>
        <datestamp>2025-01-19T09:41:49Z</datestamp>
        <setSpec>1164:5352:11553:11625</setSpec>
      </header>
      <metadata>
        <oai_dc:dc xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:oai_dc="http://www.openarchives.org/OAI/2.0/oai_dc/" xmlns="http://www.w3.org/2001/XMLSchema" xsi:schemaLocation="http://www.openarchives.org/OAI/2.0/oai_dc/ http://www.openarchives.org/OAI/2.0/oai_dc.xsd">
          <dc:title>拡散モデルを用いた異常検知に対する選択的推論</dc:title>
          <dc:title>Selective Inference for Anomaly Detection using Diﬀusion Models</dc:title>
          <dc:creator>勝岡, 輝行</dc:creator>
          <dc:creator>白石, 智洸</dc:creator>
          <dc:creator>三輪, 大貴</dc:creator>
          <dc:creator>Vo, Nguyen Le Duy</dc:creator>
          <dc:creator>竹内, 一郎</dc:creator>
          <dc:creator>Teruyuki, Katsuoka</dc:creator>
          <dc:creator>Tomohiro, Shiraishi</dc:creator>
          <dc:creator>Daiki, Miwa</dc:creator>
          <dc:creator>Vo, Nguyen Le Duy</dc:creator>
          <dc:creator>Ichiro, Takeuchi</dc:creator>
          <dc:subject>情報論的学習理論と機械学習2</dc:subject>
          <dc:description>近年，生成モデルである拡散モデルを用いた異常検知の研究が盛んに行われている．これらの異常検知手法は，正常画像のみを用いてモデルを訓練し，入力画像に対して仮想的な正常画像を生成することで，入力画像内の異常領域を同定する．しかし，このようにして得られた異常領域の信頼性については，これまで十分に議論されていなかった．この問題に対処するため，本研究では選択的推論の枠組みを用いて，異常領域の信頼性を ???? 値として定量化し，type I error rate が制御できる適切な統計的仮説検定を提案する．また，人工データ実験と実データ実験によって提案手法の妥当性を示す．</dc:description>
          <dc:description>In recent years, there has been active research on anomaly detection using diﬀusion models, which are generative models. These anomaly detection methods train the model using only normal images and generate virtual normal images for the input image to identify abnormal regions within the input. However, the reliability of the abnormal regions obtained in this way has not been suﬃciently discussed. To address this problem, we propose an appropriate statistical hypothesis testing method that quantiﬁes the reliability of the abnormal regions as a ????-value using a selective inference framework, allowing for the control of the type I error rate. The validity of the proposed method is demonstrated through experiments on synthetic and real data.</dc:description>
          <dc:description>technical report</dc:description>
          <dc:publisher>情報処理学会</dc:publisher>
          <dc:date>2024-06-13</dc:date>
          <dc:format>application/pdf</dc:format>
          <dc:identifier>研究報告バイオ情報学（BIO）</dc:identifier>
          <dc:identifier>9</dc:identifier>
          <dc:identifier>2024-BIO-78</dc:identifier>
          <dc:identifier>1</dc:identifier>
          <dc:identifier>7</dc:identifier>
          <dc:identifier>2188-8590</dc:identifier>
          <dc:identifier>AA12055912</dc:identifier>
          <dc:identifier>https://ipsj.ixsq.nii.ac.jp/record/234836/files/IPSJ-BIO24078009.pdf</dc:identifier>
          <dc:language>jpn</dc:language>
        </oai_dc:dc>
      </metadata>
    </record>
  </GetRecord>
</OAI-PMH>
