Preliminary investigation on activity recognition for packaging tasks using motif-guided attention networks

• Item type: SIG Technical Reports(1)
• 公開日: 2020-12-01
• タイトル: Preliminary investigation on activity recognition for packaging tasks using motif-guided attention networks
• タイトル / 言語: en
• タイトル: Preliminary investigation on activity recognition for packaging tasks using motif-guided attention networks
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: 行動識別
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: Department of Multimedia Engineering, Graduate School of Information Science and Technology, Osaka University
• 著者所属: Department of Multimedia Engineering, Graduate School of Information Science and Technology, Osaka University
• 著者所属: Department of Multimedia Engineering, Graduate School of Information Science and Technology, Osaka University
• 著者所属: Department of Multimedia Engineering, Graduate School of Information Science and Technology, Osaka University
• 著者所属: Corporate Manufacturing Engineering Center, Toshiba Corporation
• 著者所属: Corporate Manufacturing Engineering Center, Toshiba Corporation
• 著者所属(英): en
• 著者所属(英): Department of Multimedia Engineering, Graduate School of Information Science and Technology, Osaka University
• 著者所属(英): en
• 著者所属(英): Department of Multimedia Engineering, Graduate School of Information Science and Technology, Osaka University
• 著者所属(英): en
• 著者所属(英): Department of Multimedia Engineering, Graduate School of Information Science and Technology, Osaka University
• 著者所属(英): en
• 著者所属(英): Department of Multimedia Engineering, Graduate School of Information Science and Technology, Osaka University
• 著者所属(英): en
• 著者所属(英): Corporate Manufacturing Engineering Center, Toshiba Corporation
• 著者所属(英): en
• 著者所属(英): Corporate Manufacturing Engineering Center, Toshiba Corporation
• 著者名: Jaime, Morales; Naoya, Yoshimura; Qingxin, Xia; Takuya, Maekawa; Atsushi, Wada; Yasuo, Namioka
• 著者名(英): Jaime, Morales; Naoya, Yoshimura; Qingxin, Xia; Takuya, Maekawa; Atsushi, Wada; Yasuo, Namioka
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: This study presents a method for recognizing packaging tasks using wrist-worn accelerometer sensors under real conditions. As the lead times and actions of packaging activities depend on the number of objects to pack along with the size and shape of each object, it is difficult to recognize operations during every period. We propose a segmentation neural network augmented with a multi-head attention mechanism to capture actions found in a specific operation, which can be useful to identify individual operations. To efficiently detect useful actions with limited training data, we propose an attention guiding approach based on existing motif detection algorithms, which find actions (motifs) that frequently appear in a specific operation. We then use the occurrence of these motifs as a target for each attention head, enabling it to increase its ability to recognize similar operations during the packaging process. We evaluate our framework using data obtained in an actual logistics center.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: This study presents a method for recognizing packaging tasks using wrist-worn accelerometer sensors under real conditions. As the lead times and actions of packaging activities depend on the number of objects to pack along with the size and shape of each object, it is difficult to recognize operations during every period. We propose a segmentation neural network augmented with a multi-head attention mechanism to capture actions found in a specific operation, which can be useful to identify individual operations. To efficiently detect useful actions with limited training data, we propose an attention guiding approach based on existing motif detection algorithms, which find actions (motifs) that frequently appear in a specific operation. We then use the occurrence of these motifs as a target for each attention head, enabling it to increase its ability to recognize similar operations during the packaging process. We evaluate our framework using data obtained in an actual logistics center.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AA11838947
• 書誌情報: 研究報告ユビキタスコンピューティングシステム（UBI）; 巻 2020-UBI-68, 号 11, p. 1-8, 発行日 2020-12-01
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 2188-8698
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会