【16-03期VALSE Webinar活动】  报告嘉宾1：刘力（英国诺桑比亚大学）
报告时间：2016年1月20日（星期三）晚20:00（北京时间）
报告题目：基于特征表达和特征学习在人体行为识别中的应用
(Human Action Recognition by Feature Engineering and Feature
Learning )[Slides]
主持人： 王琦（ 西北工业大学）卢孝强（中科院西光所）
报告摘要：HUMAN action recognition, as a hot research area in computer
vision, has many potential applications such as video search and retrieval,
intelligent surveillance systems, and human-computer interaction. Despite its
popularity, how to precisely distinguish different actions still remains
challenging, since variations in lighting conditions, intra-class differences
and complex backgrounds all pose as obstacles for robust action
recognition. Generally, the basic
approach to action recognition contains the following two stages: 1) feature
extraction and representation and 2) action classification. In this talk, I
will mainly talk about the feature engineering and feature learning for action
recognition. Particularly, some discriminative hand-crafted local/global
feature embedding methods mentioned in our previous work will be reviewed
first. After that the machine-learned representations via genetic-programming
and cross-domain dictionary learning will be introduced. We demonstrate that these propose methods
have achieved significantly improvement over a variety of realistic datasets
for action recognition tasks. More details can be found in our references and
webpages.
参考文献：
[ 1 ] X.
Zhen, L. Shao and F. Zheng, “Discriminative
Embedding via Image-to-Class Distances”,
British Machine Vision Conference (BMVC), 2014. [ 2 ] L. Shao, et al., “Spatio-Temporal
Laplacian Pyramid Coding for Action Recognition”, IEEE Transactions on Cybernetics, 2014. [ 3 ] L. Liu, L. Shao and P.
Rockett, “Genetic Programming-Evolved
Spatio-Temporal Descriptor for Human Action Recognition”, British Machine Vision Conference (BMVC), Surrey, UK, 2012. [ 4 ] L. Liu and L. Shao, “Learning
Discriminative Representations from RGB-D Video Data”, International Joint Conference on Artificial Intelligence
(IJCAI), Beijing, China, 2013. [ 5 ] F. Zhu and L. Shao, “Weakly-Supervised
Cross-Domain Dictionary Learning for Visual Recognition”, International Journal of Computer Vision (IJCV), vol. 109, no.
1-2, pp. 42-59, Aug. 2014. [ 6 ] M. Yu, L. Liu and L. Shao, “Structure-Preserving Binary Representations for RGB-D Action
Recognition”, IEEE Transactions on
Pattern Analysis and Machine Intelligence (T-PAMI)，2015. [ 7 ]. Shao, L. Liu and M. Yu, “Kernelized Multiview Projection for Robust Action Recognition”, International Journal of Computer Vision (IJCV), 2015.. 报告人简介：
刘力，博士， 现任英国诺桑比亚大学计算机视觉与人工智能实验室研究员。2011年毕业于西安交通大学自动化科学系并获得学士学位。 后赴英国谢菲尔德大学电子工程系攻读博士，并于2014年 获得博士学位。2014年底加入英国诺上比亚大学工作至今。刘力博士的主要研究领域是计算机视觉，模式识别和多媒体，尤其关注人体行为识别， 辨别性特征提取及表达的监督/非监督学习， 大规模数据的检索以及深度学习在图像视频上的相关应用。相关研究成果已发表在TPAMI，TIP，TNNLS，TCYB，IJCV，
PR，ICCV，IJCAI，ACM MM 等国际顶级期刊和会议上。部分研究成果也已取得发明专利许可。目前主持并领导英国CREATEC
INNOVATION公司机器视觉相关项目的研发工作。他是IEEE会员，并担任IEEE TIP, TNNLS, TCYB、CVPR、ACM
MM等多个国际期刊和会议的审稿人。 | 