Action Recognition for Videos by Long-Term Point Trajectory Analysis with Background Removal

研究成果: 著書/レポートタイプへの貢献会議での発言

1 引用 (Scopus)

抄録

Recently, dense trajectories were shown to be an efficient video motion representation for action recognition and achieved state-of-the-art results on a variety of video datasets. This paper improves their performance by taking into account camera motion. To estimate camera motion, the authors use long-term point trajectory analysis to cluster image points and propose an algorithm to find possible background cluster from these clusters according to background nature in a video. Considering the original clusters could not segment the foreground and background very well. The authors optimize the background cluster, and use the cluster to rectify the trajectory. Experimental results on three challenging action datasets (i.e., Hollywood2, Olympic Sports and UCF50) show that the rectified trajectories significantly outperform original dense trajectories.

元の言語英語
ホスト出版物のタイトルProceedings - 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016
編集者Giuseppe De Pietro, Albert Dipanda, Richard Chbeir, Luigi Gallo, Kokou Yetongnon
出版者Institute of Electrical and Electronics Engineers Inc.
ページ23-30
ページ数8
ISBN(電子版)9781509056989
DOI
出版物ステータス出版済み - 4 21 2017
イベント12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016 - Naples, イタリア
継続期間: 11 28 201612 1 2016

その他

その他12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016
イタリア
Naples
期間11/28/1612/1/16

Fingerprint

Trajectories
Cameras
Sports
Cluster Analysis
Datasets

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition
  • Radiology Nuclear Medicine and imaging
  • Computer Networks and Communications
  • Signal Processing

これを引用

Xiang, Y., Okada, Y., & Kaneko, K. (2017). Action Recognition for Videos by Long-Term Point Trajectory Analysis with Background Removal. : G. De Pietro, A. Dipanda, R. Chbeir, L. Gallo, & K. Yetongnon (版), Proceedings - 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016 (pp. 23-30). [7907440] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SITIS.2016.13

Action Recognition for Videos by Long-Term Point Trajectory Analysis with Background Removal. / Xiang, Yu; Okada, Yoshihiro; Kaneko, Kosuke.

Proceedings - 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016. 版 / Giuseppe De Pietro; Albert Dipanda; Richard Chbeir; Luigi Gallo; Kokou Yetongnon. Institute of Electrical and Electronics Engineers Inc., 2017. p. 23-30 7907440.

研究成果: 著書/レポートタイプへの貢献会議での発言

Xiang, Y, Okada, Y & Kaneko, K 2017, Action Recognition for Videos by Long-Term Point Trajectory Analysis with Background Removal. : G De Pietro, A Dipanda, R Chbeir, L Gallo & K Yetongnon (版), Proceedings - 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016., 7907440, Institute of Electrical and Electronics Engineers Inc., pp. 23-30, 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016, Naples, イタリア, 11/28/16. https://doi.org/10.1109/SITIS.2016.13
Xiang Y, Okada Y, Kaneko K. Action Recognition for Videos by Long-Term Point Trajectory Analysis with Background Removal. : De Pietro G, Dipanda A, Chbeir R, Gallo L, Yetongnon K, 編集者, Proceedings - 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016. Institute of Electrical and Electronics Engineers Inc. 2017. p. 23-30. 7907440 https://doi.org/10.1109/SITIS.2016.13
Xiang, Yu ; Okada, Yoshihiro ; Kaneko, Kosuke. / Action Recognition for Videos by Long-Term Point Trajectory Analysis with Background Removal. Proceedings - 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016. 編集者 / Giuseppe De Pietro ; Albert Dipanda ; Richard Chbeir ; Luigi Gallo ; Kokou Yetongnon. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 23-30
@inproceedings{b2655d0c2cb6472d8d143e643cc91f82,
title = "Action Recognition for Videos by Long-Term Point Trajectory Analysis with Background Removal",
abstract = "Recently, dense trajectories were shown to be an efficient video motion representation for action recognition and achieved state-of-the-art results on a variety of video datasets. This paper improves their performance by taking into account camera motion. To estimate camera motion, the authors use long-term point trajectory analysis to cluster image points and propose an algorithm to find possible background cluster from these clusters according to background nature in a video. Considering the original clusters could not segment the foreground and background very well. The authors optimize the background cluster, and use the cluster to rectify the trajectory. Experimental results on three challenging action datasets (i.e., Hollywood2, Olympic Sports and UCF50) show that the rectified trajectories significantly outperform original dense trajectories.",
author = "Yu Xiang and Yoshihiro Okada and Kosuke Kaneko",
year = "2017",
month = "4",
day = "21",
doi = "10.1109/SITIS.2016.13",
language = "English",
pages = "23--30",
editor = "{De Pietro}, Giuseppe and Albert Dipanda and Richard Chbeir and Luigi Gallo and Kokou Yetongnon",
booktitle = "Proceedings - 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

TY - GEN

T1 - Action Recognition for Videos by Long-Term Point Trajectory Analysis with Background Removal

AU - Xiang, Yu

AU - Okada, Yoshihiro

AU - Kaneko, Kosuke

PY - 2017/4/21

Y1 - 2017/4/21

N2 - Recently, dense trajectories were shown to be an efficient video motion representation for action recognition and achieved state-of-the-art results on a variety of video datasets. This paper improves their performance by taking into account camera motion. To estimate camera motion, the authors use long-term point trajectory analysis to cluster image points and propose an algorithm to find possible background cluster from these clusters according to background nature in a video. Considering the original clusters could not segment the foreground and background very well. The authors optimize the background cluster, and use the cluster to rectify the trajectory. Experimental results on three challenging action datasets (i.e., Hollywood2, Olympic Sports and UCF50) show that the rectified trajectories significantly outperform original dense trajectories.

AB - Recently, dense trajectories were shown to be an efficient video motion representation for action recognition and achieved state-of-the-art results on a variety of video datasets. This paper improves their performance by taking into account camera motion. To estimate camera motion, the authors use long-term point trajectory analysis to cluster image points and propose an algorithm to find possible background cluster from these clusters according to background nature in a video. Considering the original clusters could not segment the foreground and background very well. The authors optimize the background cluster, and use the cluster to rectify the trajectory. Experimental results on three challenging action datasets (i.e., Hollywood2, Olympic Sports and UCF50) show that the rectified trajectories significantly outperform original dense trajectories.

UR - http://www.scopus.com/inward/record.url?scp=85019242572&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85019242572&partnerID=8YFLogxK

U2 - 10.1109/SITIS.2016.13

DO - 10.1109/SITIS.2016.13

M3 - Conference contribution

SP - 23

EP - 30

BT - Proceedings - 12th International Conference on Signal Image Technology and Internet-Based Systems, SITIS 2016

A2 - De Pietro, Giuseppe

A2 - Dipanda, Albert

A2 - Chbeir, Richard

A2 - Gallo, Luigi

A2 - Yetongnon, Kokou

PB - Institute of Electrical and Electronics Engineers Inc.

ER -