Higher-order Autoregressive Models for Dynamic Textures
british machine vision conference(2007)
摘要
Dynamic textured sequences are characterized by the interactions be- tween many particles or objects in the scene. Based on earlier work the im- ages of the sequence are interpreted as the output of a linear autoregressive process driven by white Gaussian noise. We extend earlier work by increas- ing the amount temporal information included when learning the motion in the scene, allowing the models to capture complex motion patterns which ex- tend over multiple frames, thereby increasing the perceptual accuracy of the synthesized results. To overcome problems of dynamic model stability, we apply Burg's Maximum Entropy Spectral Analysis technique f or parameter estimation, which is found to be reliably stable on smaller samples of training data, even with higher-order dynamics.
更多查看译文
关键词
autoregressive process,higher order,parameter estimation,maximum entropy,white gaussian noise,autoregressive model
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络