An important problem in speech, and generally activity, recognition is to develop analyses that are invariant to the execution rates. We introduce a theoretical framework that provides a parametrization-invariant metric for comparing parametrized paths on Riemannian manifolds. Treating instances of activities as parametrized paths on a Riemannian manifold of covariance matrices, we apply this framework to the problem of visual speech recognition from image sequences. We represent each sequence as a path on the space of covariance matrices, each covariance matrix capturing spatial variability of visual features in a frame, and perform simultaneous pairwise temporal alignment and comparison of paths. This removes the temporal variability and helps provide a robust metric for visual speech classification. We evaluated this idea on the OuluVS database and the rank-1 nearest neighbor classification rate improves from 32% to 57% due to temporal alignment.
|State||Published - 2013|
|Event||2013 4th National Conference on Computer Vision, Pattern Recognition, Image Processing and Graphics, NCVPRIPG 2013 - Jodhpur, Rajasthan, India|
Duration: Dec 18 2013 → Dec 21 2013
|Conference||2013 4th National Conference on Computer Vision, Pattern Recognition, Image Processing and Graphics, NCVPRIPG 2013|
|Period||12/18/13 → 12/21/13|