Abstract | ||
---|---|---|
We present a novel approach to non-rigid structure from motion (NRSFM) from an orthographic video sequence, based on a new interpretation of the problem. Existing ap- proaches assume the object shape space is well-modeled by a linear subspace. Our approach only assumes that small neighborhoods of shapes are well-modeled with a linear subspace. This constrains the shapes to belong to a man- ifold of dimensionality equal to the number of degrees of freedom of the object. After showing that the problem is still overconstrained, we present a solution composed of a novel initialization algorithm, followed by a robust exten- sion of the Locally Smooth Manifold Learning algorithm tailored to the NRSFM problem. We finally present some test cases where the linear basis method fails (and is ac- tually not meant to work) while the proposed approach is successful. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1109/CVPR.2008.4587679 | Anchorage, AK |
Keywords | Field | DocType |
image motion analysis,image sequences,learning (artificial intelligence),video signal processing,linear basis method,linear subspace,locally smooth manifold learning algorithm,nonrigid structure from motion,novel initialization algorithm,object shape space,orthographic video sequence | Structure from motion,Computer vision,Orthographic projection,Computer science,Curse of dimensionality,Linear subspace,Robustness (computer science),Artificial intelligence,Initialization,Nonlinear dimensionality reduction,Manifold | Conference |
Volume | Issue | ISSN |
2008 | 1 | 1063-6919 E-ISBN : 978-1-4244-2243-2 |
ISBN | Citations | PageRank |
978-1-4244-2243-2 | 37 | 1.29 |
References | Authors | |
21 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Vincent Rabaud | 1 | 270 | 14.91 |
Serge J. Belongie | 2 | 12512 | 1010.13 |