Dynamic textures synthesis as nonlinear manifold learning and traversing

Che Bin Liu, Ruei Sung Lin, Narendra Ahuja, Ming Hsuan Yang

Research output: Contribution to conferencePaper

21 Citations (Scopus)

Abstract

We formulate the problem of dynamic texture synthesis as a nonlinear manifold learning and traversing problem. We characterize dynamic textures as the temporal changes in spectral parameters of image sequences. For continuous changes of such parameters, it is commonly assumed that all these parameters lie on or close to a low-dimensional manifold embedded in the original configuration space. For complex dynamic data, the manifolds are usually nonlinear and we propose to use a mixture of linear subspaces to model a nonlinear manifold. These locally linear subspaces are further aligned within a global coordinate system. With the nonlinear manifold being globally parameterized, we overcome motion discontinuity problems encountered in switching linear models and dynamics. We present a nonparametric method to describe the complex dynamics of data sequences on the manifold. We also apply such approach to dynamic spatial parameters such as motion capture data. The experimental results suggest that our approach is able to synthesize smooth, complex dynamic textures and human motions, and has potential applications to other dynamic data synthesis problems.

Original languageEnglish
Pages859-868
Number of pages10
Publication statusPublished - 2006 Jan 1
Event2006 17th British Machine Vision Conference, BMVC 2006 - Edinburgh, United Kingdom
Duration: 2006 Sep 42006 Sep 7

Conference

Conference2006 17th British Machine Vision Conference, BMVC 2006
CountryUnited Kingdom
CityEdinburgh
Period06/9/406/9/7

Fingerprint

Textures
Data acquisition

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Cite this

Liu, C. B., Lin, R. S., Ahuja, N., & Yang, M. H. (2006). Dynamic textures synthesis as nonlinear manifold learning and traversing. 859-868. Paper presented at 2006 17th British Machine Vision Conference, BMVC 2006, Edinburgh, United Kingdom.
Liu, Che Bin ; Lin, Ruei Sung ; Ahuja, Narendra ; Yang, Ming Hsuan. / Dynamic textures synthesis as nonlinear manifold learning and traversing. Paper presented at 2006 17th British Machine Vision Conference, BMVC 2006, Edinburgh, United Kingdom.10 p.
@conference{fdcb2099502741608b687f7d3fb03f45,
title = "Dynamic textures synthesis as nonlinear manifold learning and traversing",
abstract = "We formulate the problem of dynamic texture synthesis as a nonlinear manifold learning and traversing problem. We characterize dynamic textures as the temporal changes in spectral parameters of image sequences. For continuous changes of such parameters, it is commonly assumed that all these parameters lie on or close to a low-dimensional manifold embedded in the original configuration space. For complex dynamic data, the manifolds are usually nonlinear and we propose to use a mixture of linear subspaces to model a nonlinear manifold. These locally linear subspaces are further aligned within a global coordinate system. With the nonlinear manifold being globally parameterized, we overcome motion discontinuity problems encountered in switching linear models and dynamics. We present a nonparametric method to describe the complex dynamics of data sequences on the manifold. We also apply such approach to dynamic spatial parameters such as motion capture data. The experimental results suggest that our approach is able to synthesize smooth, complex dynamic textures and human motions, and has potential applications to other dynamic data synthesis problems.",
author = "Liu, {Che Bin} and Lin, {Ruei Sung} and Narendra Ahuja and Yang, {Ming Hsuan}",
year = "2006",
month = "1",
day = "1",
language = "English",
pages = "859--868",
note = "2006 17th British Machine Vision Conference, BMVC 2006 ; Conference date: 04-09-2006 Through 07-09-2006",

}

Liu, CB, Lin, RS, Ahuja, N & Yang, MH 2006, 'Dynamic textures synthesis as nonlinear manifold learning and traversing' Paper presented at 2006 17th British Machine Vision Conference, BMVC 2006, Edinburgh, United Kingdom, 06/9/4 - 06/9/7, pp. 859-868.

Dynamic textures synthesis as nonlinear manifold learning and traversing. / Liu, Che Bin; Lin, Ruei Sung; Ahuja, Narendra; Yang, Ming Hsuan.

2006. 859-868 Paper presented at 2006 17th British Machine Vision Conference, BMVC 2006, Edinburgh, United Kingdom.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Dynamic textures synthesis as nonlinear manifold learning and traversing

AU - Liu, Che Bin

AU - Lin, Ruei Sung

AU - Ahuja, Narendra

AU - Yang, Ming Hsuan

PY - 2006/1/1

Y1 - 2006/1/1

N2 - We formulate the problem of dynamic texture synthesis as a nonlinear manifold learning and traversing problem. We characterize dynamic textures as the temporal changes in spectral parameters of image sequences. For continuous changes of such parameters, it is commonly assumed that all these parameters lie on or close to a low-dimensional manifold embedded in the original configuration space. For complex dynamic data, the manifolds are usually nonlinear and we propose to use a mixture of linear subspaces to model a nonlinear manifold. These locally linear subspaces are further aligned within a global coordinate system. With the nonlinear manifold being globally parameterized, we overcome motion discontinuity problems encountered in switching linear models and dynamics. We present a nonparametric method to describe the complex dynamics of data sequences on the manifold. We also apply such approach to dynamic spatial parameters such as motion capture data. The experimental results suggest that our approach is able to synthesize smooth, complex dynamic textures and human motions, and has potential applications to other dynamic data synthesis problems.

AB - We formulate the problem of dynamic texture synthesis as a nonlinear manifold learning and traversing problem. We characterize dynamic textures as the temporal changes in spectral parameters of image sequences. For continuous changes of such parameters, it is commonly assumed that all these parameters lie on or close to a low-dimensional manifold embedded in the original configuration space. For complex dynamic data, the manifolds are usually nonlinear and we propose to use a mixture of linear subspaces to model a nonlinear manifold. These locally linear subspaces are further aligned within a global coordinate system. With the nonlinear manifold being globally parameterized, we overcome motion discontinuity problems encountered in switching linear models and dynamics. We present a nonparametric method to describe the complex dynamics of data sequences on the manifold. We also apply such approach to dynamic spatial parameters such as motion capture data. The experimental results suggest that our approach is able to synthesize smooth, complex dynamic textures and human motions, and has potential applications to other dynamic data synthesis problems.

UR - http://www.scopus.com/inward/record.url?scp=34948858139&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34948858139&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:34948858139

SP - 859

EP - 868

ER -

Liu CB, Lin RS, Ahuja N, Yang MH. Dynamic textures synthesis as nonlinear manifold learning and traversing. 2006. Paper presented at 2006 17th British Machine Vision Conference, BMVC 2006, Edinburgh, United Kingdom.