LEARNING NONLINEAR MANIFOLDS OF DYNAMIC TEXTURES

Ishan Awasthi, Ahmed Elgammal

2006

Abstract

Dynamic textures are sequences of images of moving scenes that show stationarity properties in time. Eg: waves, flame, fountain, etc. Recent attempts at generating, potentially, infinitely long sequences model the dynamic texture as a Linear Dynamic System. This assumes a linear correlation in the input sequence. Most real world sequences however, exhibit nonlinear correlation between frames. In this paper, we propose a technique of generating dynamic textures using a low dimension model that preserves the non-linear correlation. We use nonlinear dimensionality reduction to create an embedding of the input sequence. Using this embedding, a nonlinear mapping is learnt from the embedded space into the image input space. Any input is represented by a linear combination of nonlinear bases functions centered along the manifold in the embedded space. A spline is used to move along the input manifold in this embedded space as a similar manifold is created for the output. The nonlinear mapping learnt on the input is used to map this new manifold into a sequence in the image space. Output sequences, thus created, contain images never present in the original sequence and are very realistic.

References

  1. J. Stam and E. Fiume. Depicting fire and other gaseous phenomena using diffusion processes. In Proc. SIGGRAPH 7895, pages 129-136, August 1995.
  2. J. K. Hodgins and W. L.Wooten. Animating human athletes. In Robotics Research: The Eighth International Symposium, pages 356-367, 1998.
  3. J. Popovicc, S. M. Seitz, M. Erdmann, Z. Popovíc, and A. Witkin. Interactive manipulation of rigid body simulations. In Proc. of SIGGRAPH 7800, pages 209- 218, July 2000.
  4. C. H. Perry and R. W. Picard, ”Synthesizing Flames and Their Spreading,” Proc. of the 5th Eurographics Workshop on Animation and Simulation, Oslo, Norway, Sept. 1994.
  5. A. Schödl, R. Szeliski, D.H. Salesin, I. Essa, Video textures, in: K.Akeley (Ed.), Siggraph 2000, Computer Graphics Proceedings, ACMPress/ACM SIGGRAPH/Addison Wesley/Longman, 2000, pp. 489-498.
  6. A.W. Fitzgibbon, Stochastic rigidity: image registration for nowherestatic scenes, Proceedings of the Eighth International Conference On Computer Vision, 2001, pp. 662-669.
  7. S. Soatto, G. Doretto, Y.N. Wu, Dynamic textures, International Conference on Computer Vision, 2001, pp. 439-446.
  8. V. Kwatra, A. Schodl, I. Essa, G. Turk and A. Bobick. Graphcut Textures: Image and Video Synthesis Using Graph Cuts. In Proceedings of Siggraph'03, pp. 277- 286, 2003.
  9. M. Szummer and R. W. Picard. Temporal Texture Modeling. IEEE International Conference on Image Processing, vol. 3, pp. 823-826, 1996.
  10. L. Yuan, F. Wen, C. Liu, and H.Y. shum, “Synthesizing Dynamic Texture with Closed-Loop Linear Dynamical System”, ECCV 2004, LCNS 3022, pp.603-616, 2004.
  11. S. Roweis and L. Saul, Nonlinear dimensionality reduction by locally linear embedding, Science 290 (2000), 2323{2326.
  12. J. B. Tenenbaum, V. de Silva, and J. C. Langford, A global geometric framework for nonlinear dimensionality reduction, Science 290 (2000), 2319{2323.
Download


Paper Citation


in Harvard Style

Awasthi I. and Elgammal A. (2006). LEARNING NONLINEAR MANIFOLDS OF DYNAMIC TEXTURES . In Proceedings of the First International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP, ISBN 972-8865-40-6, pages 243-250. DOI: 10.5220/0001378202430250


in Bibtex Style

@conference{visapp06,
author={Ishan Awasthi and Ahmed Elgammal},
title={LEARNING NONLINEAR MANIFOLDS OF DYNAMIC TEXTURES},
booktitle={Proceedings of the First International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP,},
year={2006},
pages={243-250},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0001378202430250},
isbn={972-8865-40-6},
}


in EndNote Style

TY - CONF
JO - Proceedings of the First International Conference on Computer Vision Theory and Applications - Volume 1: VISAPP,
TI - LEARNING NONLINEAR MANIFOLDS OF DYNAMIC TEXTURES
SN - 972-8865-40-6
AU - Awasthi I.
AU - Elgammal A.
PY - 2006
SP - 243
EP - 250
DO - 10.5220/0001378202430250