基于动作子空间和权重条件随机场的行为识别(修改稿).docVIP

基于动作子空间和权重条件随机场的行为识别 王智文1(，1,2，，1,2， (1. 广西科技大学计算机科学与通信工程学院 柳州 545006; 2. 广西信息科学实验中心) 【摘要】单目视频中的人类行为识别本文提出了基于动作子空间与权重该方法结合了基于特征提取的核主分量分析（KPCA）与基于运动建模的权重条件随机场（WCRF）模型。探讨了通过非线性降维行为空间的基本结构，并在运动轨迹投影过程中保留清晰的时间顺序，使得人体轮廓数据表示更紧凑。WCRF通过多种交互途径对时间序列建模，从而提高了信息共享的联合精确度，具有超越生成模型的优势（如放宽观察之间独立性的假设，有效地将重叠的特征和远距离依存关系合并起来的能力）。实验结果表明，提出的不仅能够准确地识别随时间、区域内外人员变化的人类行为，而且对噪声和其它因素鲁棒性强。 关 键 词 人类行为识别; WCRF; KPCA; 非线性降维; 人体轮廓提取与表示 Behavior Recognition based on Action Subspace and Weight Condition Random Field WANG Zhi-wen1,JIANG Lian-yuan1,2,OU Yang-hao1, ZHANG Can-long3,XIA Dong-xue1,HUANG Zhen-jin1,2,and WANG Peng-tao4 (1. College of computer science and communication engineering, Guangxi University of science and technology Liuzhou 545006; 2. Guangxi Experiment Center of Information Science Guilin 541004;3. School of Computer Science Information Technology, Guangxi Normal University Guilin 541004; 4. College of electrical and Information Engineering, Guangxi University of science and technology Liuzhou 545006) Abstract For human behavior recognition in monocular video, a method for recognizing human behavior based on action subspace and weighted condition random field is presented in this paper. This method combines kernel principal component analysis based on feature extraction and weighted conditional random field based on activity modeling. Silhouette data of human is represented more compactly by nonlinear dimensionality reduction that explores the basic structure of action space and preserves explicit temporal orders in the course of projection trajectories of motions. Temporal sequences are modeled in WCRF by using multiple interacting ways, thus increasing joint accuracy by information sharing, and this model has superiority over generative ones (e.g., relaxing independence assumption between observations and the ability to effectively incorporate both overlapping features and long-range dependencies). The experimental results show that the proposed behavior recognition meth