Item type |
SIG Technical Reports(1) |
公開日 |
2022-03-03 |
タイトル |
|
|
タイトル |
自己視点動作映像に対する厳密なアフォーダンスのアノテーション |
タイトル |
|
|
言語 |
en |
|
タイトル |
Precise Affordance Annotation for Egocentric Action Video Datasets |
言語 |
|
|
言語 |
eng |
キーワード |
|
|
主題Scheme |
Other |
|
主題 |
セッション6-A |
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 |
|
|
|
東京大学生産技術研究所 |
著者所属 |
|
|
|
東京大学生産技術研究所 |
著者所属 |
|
|
|
東京大学生産技術研究所 |
著者所属 |
|
|
|
東京大学生産技術研究所 |
著者所属 |
|
|
|
東京大学生産技術研究所 |
著者所属(英) |
|
|
|
en |
|
|
Institute of Industrial Science, The University of Tokyo |
著者所属(英) |
|
|
|
en |
|
|
Institute of Industrial Science, The University of Tokyo |
著者所属(英) |
|
|
|
en |
|
|
Institute of Industrial Science, The University of Tokyo |
著者所属(英) |
|
|
|
en |
|
|
Institute of Industrial Science, The University of Tokyo |
著者所属(英) |
|
|
|
en |
|
|
Institute of Industrial Science, The University of Tokyo |
著者名 |
于, 澤程
黄, 逸飛
古田, 諒佑
郷津, 優介
佐藤, 洋一
|
著者名(英) |
Zecheng, Yu
Yifei, Huang
Ryosuke, Furuta
Yusuke, Goutsu
Yoichi, Sato
|
論文抄録 |
|
|
内容記述タイプ |
Other |
|
内容記述 |
Object affordance has attracted a growing interest in computer vision. It is an important concept that builds a bridge between human ability and object property, and provides fine-grained information for other tasks like activity forecasting, scene understanding, etc. Although affordance is investigated in many previous works, existing affordance datasets failed to distinguish affordance from other concepts like action and function. In this paper, We propose an efficient affordance annotation scheme for egocentric action video datasets to address this issue, which gives a clear and accurate definition of affordance. The scheme was applied to two large-scale egocentric video datasets: EPIC-KITCHENS and HOMAGE, and tested with various benchmark tasks. |
論文抄録(英) |
|
|
内容記述タイプ |
Other |
|
内容記述 |
Object affordance has attracted a growing interest in computer vision. It is an important concept that builds a bridge between human ability and object property, and provides fine-grained information for other tasks like activity forecasting, scene understanding, etc. Although affordance is investigated in many previous works, existing affordance datasets failed to distinguish affordance from other concepts like action and function. In this paper, We propose an efficient affordance annotation scheme for egocentric action video datasets to address this issue, which gives a clear and accurate definition of affordance. The scheme was applied to two large-scale egocentric video datasets: EPIC-KITCHENS and HOMAGE, and tested with various benchmark tasks. |
書誌レコードID |
|
|
収録物識別子タイプ |
NCID |
|
収録物識別子 |
AA11131797 |
書誌情報 |
研究報告コンピュータビジョンとイメージメディア(CVIM)
巻 2022-CVIM-229,
号 39,
p. 1-6,
発行日 2022-03-03
|
ISSN |
|
|
収録物識別子タイプ |
ISSN |
|
収録物識別子 |
2188-8701 |
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 |