WEKO3
アイテム
{"_buckets": {"deposit": "55221e82-e364-449f-b701-93e356494950"}, "_deposit": {"created_by": 3, "id": "6180", "owners": [3], "pid": {"revision_id": 0, "type": "depid", "value": "6180"}, "status": "published"}, "_oai": {"id": "oai:kyutech.repo.nii.ac.jp:00006180", "sets": ["24"]}, "author_link": ["25400", "25399", "27425"], "item_21_biblio_info_6": {"attribute_name": "書誌情報", "attribute_value_mlt": [{"bibliographicIssueDates": {"bibliographicIssueDate": "2019-08-05", "bibliographicIssueDateType": "Issued"}, "bibliographicIssueNumber": "15", "bibliographicPageEnd": "3434-20", "bibliographicPageStart": "3434-1", "bibliographicVolumeNumber": "19", "bibliographic_titles": [{"bibliographic_title": "Sensors"}]}]}, "item_21_description_4": {"attribute_name": "抄録", "attribute_value_mlt": [{"subitem_description": "Labeling activity data is a central part of the design and evaluation of human activity recognition systems. The performance of the systems greatly depends on the quantity and “quality” of annotations; therefore, it is inevitable to rely on users and to keep them motivated to provide activity labels. While mobile and embedded devices are increasingly using deep learning models to infer user context, we propose to exploit on-device deep learning inference using a long short-term memory (LSTM)-based method to alleviate the labeling effort and ground truth data collection in activity recognition systems using smartphone sensors. The novel idea behind this is that estimated activities are used as feedback for motivating users to collect accurate activity labels. To enable us to perform evaluations, we conduct the experiments with two conditional methods. We compare the proposed method showing estimated activities using on-device deep learning inference with the traditional method showing sentences without estimated activities through smartphone notifications. By evaluating with the dataset gathered, the results show our proposed method has improvements in both data quality (i.e., the performance of a classification model) and data quantity (i.e., the number of data collected) that reflect our method could improve activity data collection, which can enhance human activity recognition systems. We discuss the results, limitations, challenges, and implications for on-device deep learning inference that support activity data collection. Also, we publish the preliminary dataset collected to the research community for activity recognition.", "subitem_description_type": "Abstract"}]}, "item_21_description_60": {"attribute_name": "資源タイプ", "attribute_value_mlt": [{"subitem_description": "Journal Article", "subitem_description_type": "Other"}]}, "item_21_link_62": {"attribute_name": "研究者情報", "attribute_value_mlt": [{"subitem_link_text": "https://hyokadb02.jimu.kyutech.ac.jp/html/140_ja.html", "subitem_link_url": "https://hyokadb02.jimu.kyutech.ac.jp/html/140_ja.html"}]}, "item_21_publisher_7": {"attribute_name": "出版者", "attribute_value_mlt": [{"subitem_publisher": "MDPI"}]}, "item_21_relation_12": {"attribute_name": "DOI", "attribute_value_mlt": [{"subitem_relation_type": "isIdenticalTo", "subitem_relation_type_id": {"subitem_relation_type_id_text": "info:doi/10.3390/s19153434", "subitem_relation_type_select": "DOI"}}]}, "item_21_relation_14": {"attribute_name": "情報源", "attribute_value_mlt": [{"subitem_relation_name": [{"subitem_relation_name_text": "https://doi.org/10.3390/s19153434"}], "subitem_relation_type_id": {"subitem_relation_type_id_text": "https://doi.org/10.3390/s19153434", "subitem_relation_type_select": "DOI"}}]}, "item_21_rights_13": {"attribute_name": "権利", "attribute_value_mlt": [{"subitem_rights": "Copyright (c) 2019 by the authors. Licensee MDPI, Basel, Switzerland."}, {"subitem_rights": "Creative Commons Attribution (CC BY) license"}, {"subitem_rights": "http://creativecommons.org/licenses/by/4.0/"}]}, "item_21_select_59": {"attribute_name": "査読の有無", "attribute_value_mlt": [{"subitem_select_item": "yes"}]}, "item_21_source_id_8": {"attribute_name": "ISSN", "attribute_value_mlt": [{"subitem_source_identifier": "1424-8220", "subitem_source_identifier_type": "ISSN"}]}, "item_21_subject_16": {"attribute_name": "日本十進分類法", "attribute_value_mlt": [{"subitem_subject": "548", "subitem_subject_scheme": "NDC"}]}, "item_21_text_28": {"attribute_name": "論文ID(連携)", "attribute_value_mlt": [{"subitem_text_value": "10345308"}]}, "item_21_text_36": {"attribute_name": "著者所属", "attribute_value_mlt": [{"subitem_text_value": "Graduate School of Engineering, Kyushu Institute of Technology, 1-1 Sensui-cho, Tobata-ku, Kitakyushu-shi, Fukuoka 804-8550, Japan"}, {"subitem_text_value": "Graduate School of Engineering, Kyushu Institute of Technology, 1-1 Sensui-cho, Tobata-ku, Kitakyushu-shi, Fukuoka 804-8550, Japan"}, {"subitem_text_value": "Graduate School of Engineering, Kyushu Institute of Technology, 1-1 Sensui-cho, Tobata-ku, Kitakyushu-shi, Fukuoka 804-8550, Japan"}]}, "item_21_text_63": {"attribute_name": "連携ID", "attribute_value_mlt": [{"subitem_text_value": "7883"}]}, "item_21_version_type_58": {"attribute_name": "著者版フラグ", "attribute_value_mlt": [{"subitem_version_resource": "http://purl.org/coar/version/c_970fb48d4fbd8a85", "subitem_version_type": "VoR"}]}, "item_creator": {"attribute_name": "著者", "attribute_type": "creator", "attribute_value_mlt": [{"creatorNames": [{"creatorName": "Mairittha, Nattaya"}], "nameIdentifiers": [{"nameIdentifier": "25399", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Mairittha, Tittaya"}], "nameIdentifiers": [{"nameIdentifier": "25400", "nameIdentifierScheme": "WEKO"}]}, {"creatorAffiliations": [{"affiliationNameIdentifiers": [], "affiliationNames": [{"affiliationName": "", "affiliationNameLang": "ja"}]}], "creatorNames": [{"creatorName": "Inoue, Sozo", "creatorNameLang": "en"}, {"creatorName": "井上, 創造", "creatorNameLang": "ja"}, {"creatorName": "イノウエ, ソウゾウ", "creatorNameLang": "ja-Kana"}], "familyNames": [{"familyName": "Inoue", "familyNameLang": "en"}, {"familyName": "井上", "familyNameLang": "ja"}, {"familyName": "イノウエ", "familyNameLang": "ja-Kana"}], "givenNames": [{"givenName": "Sozo", "givenNameLang": "en"}, {"givenName": "創造", "givenNameLang": "ja"}, {"givenName": "ソウゾウ", "givenNameLang": "ja-Kana"}], "nameIdentifiers": [{"nameIdentifier": "27425", "nameIdentifierScheme": "WEKO"}, {"nameIdentifier": "90346825", "nameIdentifierScheme": "e-Rad", "nameIdentifierURI": "https://nrid.nii.ac.jp/ja/nrid/1000090346825"}, {"nameIdentifier": "9335840200", "nameIdentifierScheme": "Scopus著者ID", "nameIdentifierURI": "https://www.scopus.com/authid/detail.uri?authorId=9335840200"}, {"nameIdentifier": "140", "nameIdentifierScheme": "九工大研究者情報", "nameIdentifierURI": "https://hyokadb02.jimu.kyutech.ac.jp/html/140_ja.html"}]}]}, "item_files": {"attribute_name": "ファイル情報", "attribute_type": "file", "attribute_value_mlt": [{"accessrole": "open_date", "date": [{"dateType": "Available", "dateValue": "2019-09-12"}], "displaytype": "detail", "download_preview_message": "", "file_order": 0, "filename": "s19153434.pdf", "filesize": [{"value": "2.7 MB"}], "format": "application/pdf", "future_date_message": "", "is_thumbnail": false, "licensetype": "license_free", "mimetype": "application/pdf", "size": 2700000.0, "url": {"label": "s19153434.pdf", "url": "https://kyutech.repo.nii.ac.jp/record/6180/files/s19153434.pdf"}, "version_id": "d4efd810-9ab9-4f6e-af95-e644bd582ba4"}]}, "item_language": {"attribute_name": "言語", "attribute_value_mlt": [{"subitem_language": "eng"}]}, "item_resource_type": {"attribute_name": "資源タイプ", "attribute_value_mlt": [{"resourcetype": "journal article", "resourceuri": "http://purl.org/coar/resource_type/c_6501"}]}, "item_title": "On-Device Deep Learning Inference for Efficient Activity Data Collection", "item_titles": {"attribute_name": "タイトル", "attribute_value_mlt": [{"subitem_title": "On-Device Deep Learning Inference for Efficient Activity Data Collection"}]}, "item_type_id": "21", "owner": "3", "path": ["24"], "permalink_uri": "http://hdl.handle.net/10228/00007390", "pubdate": {"attribute_name": "公開日", "attribute_value": "2019-09-12"}, "publish_date": "2019-09-12", "publish_status": "0", "recid": "6180", "relation": {}, "relation_version_is_last": true, "title": ["On-Device Deep Learning Inference for Efficient Activity Data Collection"], "weko_shared_id": 3}
On-Device Deep Learning Inference for Efficient Activity Data Collection
http://hdl.handle.net/10228/00007390
http://hdl.handle.net/10228/000073904035ae1d-4faa-4f67-9351-9800edf7cf7a
名前 / ファイル | ライセンス | アクション |
---|---|---|
s19153434.pdf (2.7 MB)
|
|
Item type | 学術雑誌論文 = Journal Article(1) | |||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
公開日 | 2019-09-12 | |||||||||||
資源タイプ | ||||||||||||
資源タイプ識別子 | http://purl.org/coar/resource_type/c_6501 | |||||||||||
資源タイプ | journal article | |||||||||||
タイトル | ||||||||||||
タイトル | On-Device Deep Learning Inference for Efficient Activity Data Collection | |||||||||||
言語 | ||||||||||||
言語 | eng | |||||||||||
著者 |
Mairittha, Nattaya
× Mairittha, Nattaya× Mairittha, Tittaya× 井上, 創造
WEKO
27425
|
|||||||||||
抄録 | ||||||||||||
内容記述タイプ | Abstract | |||||||||||
内容記述 | Labeling activity data is a central part of the design and evaluation of human activity recognition systems. The performance of the systems greatly depends on the quantity and “quality” of annotations; therefore, it is inevitable to rely on users and to keep them motivated to provide activity labels. While mobile and embedded devices are increasingly using deep learning models to infer user context, we propose to exploit on-device deep learning inference using a long short-term memory (LSTM)-based method to alleviate the labeling effort and ground truth data collection in activity recognition systems using smartphone sensors. The novel idea behind this is that estimated activities are used as feedback for motivating users to collect accurate activity labels. To enable us to perform evaluations, we conduct the experiments with two conditional methods. We compare the proposed method showing estimated activities using on-device deep learning inference with the traditional method showing sentences without estimated activities through smartphone notifications. By evaluating with the dataset gathered, the results show our proposed method has improvements in both data quality (i.e., the performance of a classification model) and data quantity (i.e., the number of data collected) that reflect our method could improve activity data collection, which can enhance human activity recognition systems. We discuss the results, limitations, challenges, and implications for on-device deep learning inference that support activity data collection. Also, we publish the preliminary dataset collected to the research community for activity recognition. | |||||||||||
書誌情報 |
Sensors 巻 19, 号 15, p. 3434-1-3434-20, 発行日 2019-08-05 |
|||||||||||
出版社 | ||||||||||||
出版者 | MDPI | |||||||||||
DOI | ||||||||||||
関連タイプ | isIdenticalTo | |||||||||||
識別子タイプ | DOI | |||||||||||
関連識別子 | info:doi/10.3390/s19153434 | |||||||||||
日本十進分類法 | ||||||||||||
主題Scheme | NDC | |||||||||||
主題 | 548 | |||||||||||
ISSN | ||||||||||||
収録物識別子タイプ | ISSN | |||||||||||
収録物識別子 | 1424-8220 | |||||||||||
著作権関連情報 | ||||||||||||
権利情報 | Copyright (c) 2019 by the authors. Licensee MDPI, Basel, Switzerland. | |||||||||||
著作権関連情報 | ||||||||||||
権利情報 | Creative Commons Attribution (CC BY) license | |||||||||||
著作権関連情報 | ||||||||||||
権利情報 | http://creativecommons.org/licenses/by/4.0/ | |||||||||||
出版タイプ | ||||||||||||
出版タイプ | VoR | |||||||||||
出版タイプResource | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |||||||||||
査読の有無 | ||||||||||||
値 | yes | |||||||||||
研究者情報 | ||||||||||||
https://hyokadb02.jimu.kyutech.ac.jp/html/140_ja.html | ||||||||||||
論文ID(連携) | ||||||||||||
10345308 | ||||||||||||
連携ID | ||||||||||||
7883 | ||||||||||||
資料タイプ | ||||||||||||
内容記述タイプ | Other | |||||||||||
内容記述 | Journal Article | |||||||||||
著者所属 | ||||||||||||
Graduate School of Engineering, Kyushu Institute of Technology, 1-1 Sensui-cho, Tobata-ku, Kitakyushu-shi, Fukuoka 804-8550, Japan | ||||||||||||
著者所属 | ||||||||||||
Graduate School of Engineering, Kyushu Institute of Technology, 1-1 Sensui-cho, Tobata-ku, Kitakyushu-shi, Fukuoka 804-8550, Japan | ||||||||||||
著者所属 | ||||||||||||
Graduate School of Engineering, Kyushu Institute of Technology, 1-1 Sensui-cho, Tobata-ku, Kitakyushu-shi, Fukuoka 804-8550, Japan | ||||||||||||
情報源 | ||||||||||||
識別子タイプ | DOI | |||||||||||
関連識別子 | https://doi.org/10.3390/s19153434 | |||||||||||
関連名称 | https://doi.org/10.3390/s19153434 |