{"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00209834","sets":["1164:4619:10416:10532"]},"path":["10532"],"owner":"44499","recid":"209834","title":["不審物検知におけるMixupの適用及びU-Netの改良に関する検討"],"pubdate":{"attribute_name":"公開日","attribute_value":"2021-02-25"},"_buckets":{"deposit":"2ef4aad1-b542-4e39-b981-4effc850005b"},"_deposit":{"id":"209834","pid":{"type":"depid","value":"209834","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"不審物検知におけるMixupの適用及びU-Netの改良に関する検討","author_link":["530070","530073","530075","530071","530068","530077","530072","530069","530076","530074"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"不審物検知におけるMixupの適用及びU-Netの改良に関する検討"},{"subitem_title":"A Consideration on Suspicious Object Detection by Mixup and Improved U-Net","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"セッション5-1","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2021-02-25","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"早稲田大学大学院基幹理工学研究科情報理工・情報通信専攻"},{"subitem_text_value":"早稲田大学理工学術院"},{"subitem_text_value":"早稲田大学理工学術院"},{"subitem_text_value":"早稲田大学理工学術院"},{"subitem_text_value":"早稲田大学理工学術院"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Dept. of Comp. Sc. and Communications Eng.,Waseda Univ.","subitem_text_language":"en"},{"subitem_text_value":"Dept. of Comp. Sc. and Communications Eng.,Waseda Univ.","subitem_text_language":"en"},{"subitem_text_value":"Dept. of Comp. Sc. and Communications Eng.,Waseda Univ.","subitem_text_language":"en"},{"subitem_text_value":"Dept. of Comp. Sc. and Communications Eng.,Waseda Univ.","subitem_text_language":"en"},{"subitem_text_value":"Dept. of Comp. Sc. and Communications Eng.,Waseda Univ.","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/209834/files/IPSJ-CVIM21225036.pdf","label":"IPSJ-CVIM21225036.pdf"},"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-CVIM21225036.pdf","filesize":[{"value":"2.2 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"20"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_login","version_id":"8fbeca9c-81df-48fb-a46a-d782ea6a63d5","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2021 by the Institute of Electronics, Information and Communication Engineers This SIG report is only available to those in membership of the SIG."}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"菅野, 成希"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"亀山, 渉"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"佐藤, 俊雄"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"勝山, 裕"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"佐藤, 拓朗"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Naruki, Kanno","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Wataru, Kameyama","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Toshio, Sato","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Yutaka, Katsuyama","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Takuro, Sato","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11131797","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8701","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"本稿では，セマンティックセグメンテーションを利用した不審物検知において，Mixup によるデータオーギュメンテーションの有効性を検討し，また，U-Net を改良したモデルを提案してその精度の比較を行う．対象とする画像は W 帯を利用したパッシブイメージング技術により生成される画像である．検知対象となる物体は，背景，人物，不審物 6 種類（刃物，銃，携帯電話，液体，粉体，模擬爆弾），ファントムの計 9 種類である．取得した画像 1,008 枚に対して人手によりアノーテーション画像を生成し，これを正解とした．Mixup によるデータオーギュメンテーションの検討では，U-Net を用い，水平反転とスケール変換を合わせるデータオーギュメンテーションと比較したところ，Mixup を用いる場合の mIoU（Mean Intersection over Union）が 8.4 ポイント高く，86.0% であった．U-Net の改良モデルの比較では，Mixup を使用して，通常の U-Net，エンコーダとデコーダにそれぞれ RB（Residual Block）を適用した U-Net，両者に RB を適用した U-Net，エンコーダとデコーダにそれぞれ DB（Dense Block）を適用した U-Net，FC-DenseNet の計 7 個のモデルを比較した結果，デコーダにのみ RB を適用した U-Net が最も mIoU が高く，86.9％ であった．これらから，本実験画像においての Mixup によるデータオーギュメンテーションの有効性を確認し，また，実験画像中のノイズの影響を軽減するために，デコーダに RB を適用するのが有効であることが示唆された．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"In this paper, on suspicious object detection by using semantic segmentation, we study the effectiveness of Mixup data augmentation, propose several improved U-Net models, and compare them by precision. The target images are obtained by passive-imaging technology using W band. The objects to be detected are 9 classes including background, human, 6 types of suspicious objects (blade, gun, cell phone, liquid, powder, dummy bomb), and phantom. Annotation images are manually generated for the obtained 1,008 images, and taken as ground truth. On Mixup data augmentation, we compare it with conventional data augmentation methods including horizontal flip and scale augmentation by using U-Net, where mIoU (Mean Intersection over Union) of Mixup achieves 86.0% accuracy which is 8.4 point higher than others. On improved U-Net, we compare 7 models with Mixup, including normal U-Net, U-Net applying RB (Residual Block) to each encoder and decoder, U-Net applying RB to both of encoder and decoder, U-Net applying dense block to each encoder and decoder, and FC-DenseNet. The experimental results show that mIoU of U-Net applying RB only to decoder achieves the highest accuracy of 86.9%. These results suggest the effectiveness of Mixup data augmentation for the experimental images and applying RB for decoder to U-Net in order to reduce the noise in the experimental images.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告コンピュータビジョンとイメージメディア（CVIM）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2021-02-25","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"36","bibliographicVolumeNumber":"2021-CVIM-225"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":209834,"updated":"2025-01-19T18:22:32.038111+00:00","links":{},"created":"2025-01-19T01:11:07.110544+00:00"}