MODELING DYNAMIC TEXTURES USING SUBSPACE MIXTURES (ThuAmOR5)

Author(s) : Che-Bin Liu (University of Illinois at Urbana-Champaign, United States of America) Ruei-sung Lin (University of Illinois at Urbana-Champaign, United States of America) Narendra Ahuja (University of Illinois at Urbana-Champaign, United States of America)

Abstract : In this paper, we aim at modeling video sequences that exhibit temporal appearance variation. The dynamic texture model proposed by Soatto is effective to model simple dynamic scenes. However, because of its over-simplified appearance model and under-constrained dynamics model, the visual quality of its synthesized video sequences is often not satisfactory. This leads to our new model: we parameterize the nonlinear image manifold using mixtures of probabilistic principal component analyzers. We then align coefficients from different mixture components in a global coordinate system, and model the image dynamics in the global coordinate using an autoregressive (AR) process. The experimental results show that our method is capable of capturing complex temporal appearance variation and offers improved synthesis results over previous works.

Menu