Dynamic textures

被引:637
作者
Doretto, G [1 ]
Chiuso, A
Wu, YN
Soatto, S
机构
[1] Univ Calif Los Angeles, Dept Comp Sci, Los Angeles, CA 90095 USA
[2] Univ Padua, Dipartimento Ingn Informaz, I-35131 Padua, Italy
[3] Univ Calif Los Angeles, Dept Stat, Los Angeles, CA 90095 USA
关键词
textures; dynamic scene analysis; 3D textures; minimum description length; image compression; generative model; prediction error methods; ARMA model; subspace system identification; canonical correlation; learning;
D O I
10.1023/A:1021669406132
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Dynamic textures are sequences of images of moving scenes that exhibit certain stationarity properties in time; these include sea-waves, smoke, foliage, whirlwind etc. We present a characterization of dynamic textures that poses the problems of modeling, learning, recognizing and synthesizing dynamic textures on a firm analytical footing. We borrow tools from system identification to capture the "essence" of dynamic textures; we do so by learning (i.e. identifying) models that are optimal in the sense of maximum likelihood or minimum prediction error variance. For the special case of second-order stationary processes, we identify the model sub-optimally in closed-form. Once learned, a model has predictive power and can be used for extrapolating synthetic sequences to infinite length with negligible computational cost. We present experimental evidence that, within our framework, even low-dimensional models can capture very complex visual phenomena.
引用
收藏
页码:91 / 109
页数:19
相关论文
共 48 条
[1]   Blind source separation - Semiparametric statistical approach [J].
Amari, S ;
Cardoso, JF .
IEEE TRANSACTIONS ON SIGNAL PROCESSING, 1997, 45 (11) :2692-2700
[2]   BALANCED APPROXIMATION OF STOCHASTIC-SYSTEMS [J].
ARUN, KS ;
KUNG, SY .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 1990, 11 (01) :42-68
[3]   Texture mixing and texture movie synthesis using statistical learning [J].
Bar-Joseph, Z ;
El-Yaniv, R ;
Lischinski, D ;
Werman, M .
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS, 2001, 7 (02) :120-135
[4]  
Barzel Ronen., 1992, PHYSICALLY BASED MOD
[5]   Consistency and asymptotic normality of some subspace algorithms for systems without observed inputs [J].
Bauer, D ;
Deistler, M ;
Scherrer, W .
AUTOMATICA, 1999, 35 (07) :1243-1254
[6]   N-FOLDED SYMMETRIES BY COMPLEX MOMENTS IN GABOR SPACE AND THEIR APPLICATION TO UNSUPERVISED TEXTURE SEGMENTATION [J].
BIGUN, J ;
DUBUF, JM .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1994, 16 (01) :80-87
[7]  
BISSACCO A, 2001, P IEEE COMP SOC C CO
[8]  
BLAKE A., 1998, Active Contours
[9]  
CHIUSO A, IN PRESS J ECONOMETR
[10]   MARKOV RANDOM FIELD TEXTURE MODELS [J].
CROSS, GR ;
JAIN, AK .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1983, 5 (01) :25-39