[1]叶果,程洪,赵洋.电影中吸烟活动识别[J].智能系统学报,2011,6(5):440-444.
YE Guo,CHENG Hong,ZHAO Yang.moking recognition in movies[J].CAAI Transactions on Intelligent Systems,2011,6(5):440-444.
点击复制
YE Guo,CHENG Hong,ZHAO Yang.moking recognition in movies[J].CAAI Transactions on Intelligent Systems,2011,6(5):440-444.
电影中吸烟活动识别
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
6
期数:
2011年第5期
页码:
440-444
栏目:
学术论文—机器感知与模式识别
出版日期:
2011-10-30
- Title:
- moking recognition in movies
- 文章编号:
- 1673-4785(2011)05-0440-05
- 关键词:
- 电影; 吸烟活动识别; 纯贝叶斯互信息最大化; 计算机视觉; 模式识别
- Keywords:
- movies; smoking action recognition; naiveBayesian mutual information maximization; computer vision; pattern recognition
- 分类号:
- TP391.4
- 文献标志码:
- A
- 摘要:
- 电影中的活动识别是计算机视觉领域的一个难点问题.传统识别算法受到电影中镜头视角变化、场景变化和光照变化等因素的影响，使得其对于真实场景活动识别的效果较差.针对上述问题，提出一种新颖的基于互信息的组合识别方法.该方法以纯贝叶斯互信息最大化构造初始框架，针对“吸烟”这类极具代表性的动作，将活动的SIFT信息和STIP信息融合得到最优的组合分类器.该方法在电影《咖啡和烟》中进行了测试，实验结果表明，该方法具有很好的鲁棒性，并且很大程度上提高了抽烟活动的识别率.
- Abstract:
- Action recognition in movies is a difficult problem in the computer vision domain. Traditional approaches have a bad recognition effect because they are subjected to viewpoint changes, scene changes, and illumination changes in real scenes. This paper presented a novel combined recognition approach, using mutual information to solve the problems mentioned above. This method builds the initial skeleton using naiveBayesian mutual information maximization (NBMIM) and combines the shape information with the motion information to recognize smoking, which is a typical activity in movies. The proposed smoking recognition approach was evaluated in the film 〖WTBX〗Coffee and Cigarettes〖WTBZ〗. The results indicate that the proposed method is robust, and it significantly improves the recognition rate.
- 参考文献/References:
-
［1］LAPTEV I, MARSZALEK M, SCHMID C, et al. Learning realistic human actions from movies［C］//Proceedings of CVPR: IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, USA, 2008: 18.
?［2］GAIDON A, MARSZALEK M, SCHMID C. Mining visual actions from movies［C］//Proceedings of BMVC: British Machine Vision Conference. London, UK, 2009: 111.
［3］WANG J Z, GEMAN D, LUO Jiebo, et al. Realworld image annotation and retrieval: an introduction to the special section［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008, 30(11): 18731876.
［4］LAPTEV I. On spacetime interest points［J］. International Journal of Computer Vision, 2005, 64(2/3): 107123.
［5］LOWE D G. Distinctive image features from scaleinvariant keypoints［J］. International Journal of Computer Vision, 2004, 60(2): 91110.
［6］ALI S, BASHARAT A, SHAH M. Chaotic invariants for human action recognition［C］//Proceedings of ICCV: IEEE International Conference on Computer Vision. Rio de Janeiro, Brazil, 2007: 1421.
［7］NGUYEN N T, PHUNG D Q, VENKATESH S, et al. Learning and detecting activities from movement trajectories using the hierarchical hidden Markov models［C］//Proceedings of CVPR: IEEE Conference on Computer Vision and Pattern Recognition. San Diego, USA， 2005: 955960.
［8］MOESLUND T B, HILTON A, KRUGER V. A survey of advances in visionbased human motion capture and analysis［J］. Computer Vision and Image Understanding, 2006, 104(2): 90126.
［9］DUAN Lixin, XU Dong, TSANG I W, et al. Visual event recognition in videos by learning from web data［C］//Proceedings of CVPR: IEEE Conference on Computer Vision and Pattern Recognition. San Francisco, USA, 2010: 19591966.
［10］CAO Liangliang, LIU Zicheng, HUANG T. Crossdataset action detection［C］//Proceedings of CVPR: IEEE Conference on Computer Vision and Pattern Recognition. San Francisco, USA, 2010: 19982005.
［11］NATARAJAN P, NEVATIA R. View and scale invariant action recognition using multiview shapeflow models［C］//Proceedings of CVPR: IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, USA, 2008: 18.
［12］VITALADEVUNI S N, KELLOKUMPU V, DAVIS L S. Action recognition using ballistic dynamics［C］//Proceedings of CVPR: IEEE Conference on Computer Vision and Pattern Recognition. Anchorage, USA, 2008: 18.
［13］YILMAZ A, SHAH M. Actions sketch: a novel action representation［C］//Proceedings of CVPR: IEEE Conference on Computer Vision and Pattern Recognition. San Diego, USA, 2005: 984989.
［14］YUAN Junsong, LIU Zicheng, WU Ying. Discriminative subvolume search for efficient action detection［C］//Proceedings of CVPR: IEEE Conference on Computer Vision and Pattern Recognition. Miami, USA, 2009: 24422449.
［15］LAPTEV I, PEREZ P. Retrieving actions in movies［C］//Proceedings of ICCV: IEEE International Conference on Computer Vision. Rio de Janeiro, Brazil, 2007: 18.
［16］WU Pin, HSIEH J H, CHENG J C, et al. Human smoking event detection using visual interaction clues［C］//Proceedings of ICPR: IEEE International Conference on Pattern Recognition. Istanbul, Turkey, 2010: 43344347.
备注/Memo
收稿日期：2011-03-29.
基金项目：国家“973”计划资助项目（2011CB707000）；国家自然科学基金资助项目（61075045）；中央高校基本科研业务费专项基金资助项目（ZYGX2009X013）.
通信作者叶果.E-mail:yeguo0112@gmail.com.
作者简介：
叶果，男，1990年生，本科生，主要研究方向为人的活动识别、计算机视觉与模式识别.
程洪，男，1973年生，教授，博士生导师，博士，IEEE和ACM会员，2006—2009年在美国卡内基梅隆大学计算机学院进行博士后研究.主要研究方向为机器人、计算机视觉与模式识别、机器学习.先后主持和参与包括国家“973”计划项目、国家“863”计划项目，以及重要企业横向项目等10余项科研项目.发表学术论文40余篇，出版教材和专著各1部.
赵洋，男，1988年生，硕士研究生，主要研究方向为计算机视觉与模式识别.
更新日期/Last Update:
2011-11-16