講演名 | 2017-12-17 Action Sequence Recognition in Videos by Combining a CTC Network with a Statistical Language Model Mengxi Lin(東工大), Nakamasa Inoue(東工大), Koichi Shinoda(東工大), |
---|---|
PDFダウンロードページ | PDFダウンロードページへ |
抄録(和) | Action sequence recognition aims to recognize what actions occur in a video and their temporal order. In this paper, we propose to combine an LSTM network trained with Connectionist Temporal Classification (CTC) with a statistical language model for action sequence recognition. The statistical language model captures the relations between action instances, which are hardly learned by the CTC network. Our experiments on the Breakfast dataset show that the statistical language model can significantly boost the recognition accuracy of the CTC network, from 37.0% to 43.4%. |
抄録(英) | Action sequence recognition aims to recognize what actions occur in a video and their temporal order. In this paper, we propose to combine an LSTM network trained with Connectionist Temporal Classification (CTC) with a statistical language model for action sequence recognition. The statistical language model captures the relations between action instances, which are hardly learned by the CTC network. Our experiments on the Breakfast dataset show that the statistical language model can significantly boost the recognition accuracy of the CTC network, from 37.0% to 43.4%. |
キーワード(和) | connectionist temporal classification / action sequence recognition / statistical language model / weakly supervised learning |
キーワード(英) | connectionist temporal classification / action sequence recognition / statistical language model / weakly supervised learning |
資料番号 | PRMU2017-101 |
発行日 | 2017-12-10 (PRMU) |
研究会情報 | |
研究会 | PRMU |
---|---|
開催期間 | 2017/12/16(から2日開催) |
開催地(和) | 慶應義塾大学 矢上キャンパス |
開催地(英) | |
テーマ(和) | PRMUグランドチャレンジとコンピュータービジョン勉強会 |
テーマ(英) | |
委員長氏名(和) | 佐藤 真一(NII) |
委員長氏名(英) | Shinichi Sato(NII) |
副委員長氏名(和) | 藤吉 弘亘(中部大) / 井尻 善久(オムロン) |
副委員長氏名(英) | Hironobu Fujiyoshi(Chubu Univ.) / Yoshihisa Ijiri(Omron) |
幹事氏名(和) | 大西 正輝(産総研) / 舩冨 卓哉(奈良先端大) |
幹事氏名(英) | Masaki Oonishi(AIST) / Takuya Funatomi(NAIST) |
幹事補佐氏名(和) | 石井 雅人(NEC) / 菅野 裕介(阪大) |
幹事補佐氏名(英) | Masato Ishii(NEC) / Yusuke Sugano(Osaka Univ.) |
講演論文情報詳細 | |
申込み研究会 | Technical Committee on Pattern Recognition and Media Understanding |
---|---|
本文の言語 | ENG |
タイトル(和) | |
サブタイトル(和) | |
タイトル(英) | Action Sequence Recognition in Videos by Combining a CTC Network with a Statistical Language Model |
サブタイトル(和) | |
キーワード(1)(和/英) | connectionist temporal classification / connectionist temporal classification |
キーワード(2)(和/英) | action sequence recognition / action sequence recognition |
キーワード(3)(和/英) | statistical language model / statistical language model |
キーワード(4)(和/英) | weakly supervised learning / weakly supervised learning |
第 1 著者 氏名(和/英) | Mengxi Lin / Mengxi Lin |
第 1 著者 所属(和/英) | Tokyo Institute of Technology(略称:東工大) Tokyo Institute of Technology(略称:Tokyo Tech) |
第 2 著者 氏名(和/英) | Nakamasa Inoue / Nakamasa Inoue |
第 2 著者 所属(和/英) | Tokyo Institute of Technology(略称:東工大) Tokyo Institute of Technology(略称:Tokyo Tech) |
第 3 著者 氏名(和/英) | Koichi Shinoda / Koichi Shinoda |
第 3 著者 所属(和/英) | Tokyo Institute of Technology(略称:東工大) Tokyo Institute of Technology(略称:Tokyo Tech) |
発表年月日 | 2017-12-17 |
資料番号 | PRMU2017-101 |
巻番号(vol) | vol.117 |
号番号(no) | PRMU-362 |
ページ範囲 | pp.1-6(PRMU), |
ページ数 | 6 |
発行日 | 2017-12-10 (PRMU) |