WEKO3
アイテム
{"_buckets": {"deposit": "0b5ca7c5-de3f-4822-a38a-997eddc357b1"}, "_deposit": {"created_by": 29, "id": "1402", "owners": [29], "pid": {"revision_id": 0, "type": "depid", "value": "1402"}, "status": "published"}, "_oai": {"id": "oai:oist.repo.nii.ac.jp:00001402", "sets": ["203"]}, "author_link": ["8494", "8492", "8493", "8491"], "item_10003_biblio_info_7": {"attribute_name": "書誌情報", "attribute_value_mlt": [{"bibliographicIssueDates": {"bibliographicIssueDate": "2018-10-24", "bibliographicIssueDateType": "Issued"}, "bibliographicPageEnd": "135", "bibliographicPageStart": "134", "bibliographic_titles": [{"bibliographic_title": "第 28 回 日本神経回路学会全国大会 講演論文集"}, {"bibliographic_title": "The Proceedings of the 28th Annual Conference of the Japanese Neural Network Society ", "bibliographic_titleLang": "en"}]}]}, "item_10003_description_5": {"attribute_name": "抄録", "attribute_value_mlt": [{"subitem_description": "The ability to plan and visualize object manipulation in advance is vital for both humans and robots to smoothly reach a desired goal state. In this work, we demonstrate how our predictive coding based deep visuomotor recurrent neural network (PDVMRNN) can generate plans for a robot to manipulate objects based on a visual goal. A Tokyo Robotics Torobo Arm robot and a basic USB camera were used to record visuo-proprioceptive sequences of object manipulation. Although limitations in resolution resulted in lower success rates when plans were executed with the robot, our model is able to generate long predictions from novel start and goal states based on the learned patterns.", "subitem_description_type": "Abstract"}]}, "item_10003_publisher_8": {"attribute_name": "出版者", "attribute_value_mlt": [{"subitem_publisher": "Japanese Neural Network Society"}]}, "item_10003_relation_17": {"attribute_name": "関連サイト", "attribute_value_mlt": [{"subitem_relation_type_id": {"subitem_relation_type_id_text": "http://jnns.org/conference/2018/en/program.html ", "subitem_relation_type_select": "URI"}}]}, "item_10003_rights_15": {"attribute_name": "権利", "attribute_value_mlt": [{"subitem_rights": "© 2018 Japanese Neural Network Society"}]}, "item_10003_version_type_20": {"attribute_name": "著者版フラグ", "attribute_value_mlt": [{"subitem_version_resource": "http://purl.org/coar/version/c_ab4af688f83e57aa", "subitem_version_type": "AM"}]}, "item_creator": {"attribute_name": "著者", "attribute_type": "creator", "attribute_value_mlt": [{"creatorNames": [{"creatorName": "Matsumoto, Takazumi", "creatorNameLang": "en"}], "nameIdentifiers": [{"nameIdentifier": "8491", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Choi, Minkyu ", "creatorNameLang": "en"}], "nameIdentifiers": [{"nameIdentifier": "8492", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Jung, Minju", "creatorNameLang": "en"}], "nameIdentifiers": [{"nameIdentifier": "8493", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Tani, Jun", "creatorNameLang": "en"}], "nameIdentifiers": [{"nameIdentifier": "8494", "nameIdentifierScheme": "WEKO"}]}]}, "item_files": {"attribute_name": "ファイル情報", "attribute_type": "file", "attribute_value_mlt": [{"accessrole": "open_date", "date": [{"dateType": "Available", "dateValue": "2020-04-24"}], "displaytype": "detail", "download_preview_message": "", "file_order": 0, "filename": "PDVMRNN Matsumoto.pdf", "filesize": [{"value": "377.1 kB"}], "format": "application/pdf", "future_date_message": "", "is_thumbnail": false, "licensetype": "license_free", "mimetype": "application/pdf", "size": 377100.0, "url": {"label": "PDVMRNN Matsumoto", "url": "https://oist.repo.nii.ac.jp/record/1402/files/PDVMRNN Matsumoto.pdf"}, "version_id": "41765f85-d77b-4ef9-90ea-51a8af7ee578"}]}, "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": "conference paper", "resourceuri": "http://purl.org/coar/resource_type/c_5794"}]}, "item_title": "Generating Goal-directed Visuomotor Plans with Supervised Learning using a Predictive Coding Deep Visuomotor Recurrent Neural Network", "item_titles": {"attribute_name": "タイトル", "attribute_value_mlt": [{"subitem_title": "Generating Goal-directed Visuomotor Plans with Supervised Learning using a Predictive Coding Deep Visuomotor Recurrent Neural Network", "subitem_title_language": "en"}]}, "item_type_id": "10003", "owner": "29", "path": ["203"], "permalink_uri": "https://oist.repo.nii.ac.jp/records/1402", "pubdate": {"attribute_name": "公開日", "attribute_value": "2020-04-24"}, "publish_date": "2020-04-24", "publish_status": "0", "recid": "1402", "relation": {}, "relation_version_is_last": true, "title": ["Generating Goal-directed Visuomotor Plans with Supervised Learning using a Predictive Coding Deep Visuomotor Recurrent Neural Network"], "weko_shared_id": 29}
Generating Goal-directed Visuomotor Plans with Supervised Learning using a Predictive Coding Deep Visuomotor Recurrent Neural Network
https://oist.repo.nii.ac.jp/records/1402
https://oist.repo.nii.ac.jp/records/14020cda3489-3ced-4f27-b07a-94e5fa1ed419
名前 / ファイル | ライセンス | アクション |
---|---|---|
PDVMRNN Matsumoto (377.1 kB)
|
|
Item type | 会議発表論文 / Conference Paper(1) | |||||
---|---|---|---|---|---|---|
公開日 | 2020-04-24 | |||||
タイトル | ||||||
言語 | en | |||||
タイトル | Generating Goal-directed Visuomotor Plans with Supervised Learning using a Predictive Coding Deep Visuomotor Recurrent Neural Network | |||||
言語 | ||||||
言語 | eng | |||||
資源タイプ | ||||||
資源タイプ識別子 | http://purl.org/coar/resource_type/c_5794 | |||||
資源タイプ | conference paper | |||||
著者 |
Matsumoto, Takazumi
× Matsumoto, Takazumi× Choi, Minkyu× Jung, Minju× Tani, Jun |
|||||
書誌情報 |
第 28 回 日本神経回路学会全国大会 講演論文集 en : The Proceedings of the 28th Annual Conference of the Japanese Neural Network Society p. 134-135, 発行日 2018-10-24 |
|||||
抄録 | ||||||
内容記述タイプ | Abstract | |||||
内容記述 | The ability to plan and visualize object manipulation in advance is vital for both humans and robots to smoothly reach a desired goal state. In this work, we demonstrate how our predictive coding based deep visuomotor recurrent neural network (PDVMRNN) can generate plans for a robot to manipulate objects based on a visual goal. A Tokyo Robotics Torobo Arm robot and a basic USB camera were used to record visuo-proprioceptive sequences of object manipulation. Although limitations in resolution resulted in lower success rates when plans were executed with the robot, our model is able to generate long predictions from novel start and goal states based on the learned patterns. | |||||
出版者 | ||||||
出版者 | Japanese Neural Network Society | |||||
権利 | ||||||
権利情報 | © 2018 Japanese Neural Network Society | |||||
関連サイト | ||||||
識別子タイプ | URI | |||||
関連識別子 | http://jnns.org/conference/2018/en/program.html | |||||
著者版フラグ | ||||||
出版タイプ | AM | |||||
出版タイプResource | http://purl.org/coar/version/c_ab4af688f83e57aa |