Name
Title | ||
|---|---|---|
Rate-Invariant Analysis of Trajectories on Riemannian Manifolds with Application in Visual Speech Recognition |
Abstract | ||
|---|---|---|
In statistical analysis of video sequences for speech recognition, and more generally activity recognition, it is natural to treat temporal evolutions of features as trajectories on Riemannian manifolds. However, different evolution patterns result in arbitrary parameterizations of these trajectories. We investigate a recent framework from statistics literature that handles this nuisance variability using a cost function/distance for temporal registration and statistical summarization & modeling of trajectories. It is based on a mathematical representation of trajectories, termed transported square-root vector field (TSRVF), and the L2 norm on the space of TSRVFs. We apply this framework to the problem of speech recognition using both audio and visual components. In each case, we extract features, form trajectories on corresponding manifolds, and compute parametrization-invariant distances using TSRVFs for speech classification. On the OuluVS database the classification performance under metric increases significantly, by nearly 100% under both modalities and for all choices of features. We obtained speaker-dependent classification rate of 70% and 96% for visual and audio components, respectively. |
Year | DOI | Venue |
|---|---|---|
2014 | 10.1109/CVPR.2014.86 | CVPR |
Keywords | Field | DocType |
riemannian manifolds,visual speech recognition,statistical summarization,l2 norm,speech recognition,video sequences,speaker-dependent classification rate,statistical analysis,temporal registration,speech classification,rate-invariant analysis,audio components,transported square-root vector field,trajectories modeling,parametrization-invariant distances,feature extraction,image sequences,tsrvf,ouluvs database,temporal evolutions,visual components,arbitrary parameterizations,evolution patterns,vectors | Computer science,Artificial intelligence,Representation (mathematics),Manifold,Automatic summarization,Computer vision,Activity recognition,Pattern recognition,Vector field,Speech recognition,Invariant (mathematics),Norm (mathematics),Classification rate | Conference |
Volume | Issue | ISSN |
2014 | 1 | 1063-6919 |
Citations | PageRank | References |
19 | 0.71 | 14 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Jing-yong Su | 1 | 156 | 10.93 |
| Anuj Srivastava | 2 | 2853 | 199.47 |
| Fillipe Dias Moreira de Souza | 3 | 55 | 5.81 |
| Sudeep Sarkar | 4 | 2839 | 317.68 |