Abstract

In this paper, a novel approach is presented for motion retrieval based on reference index that reduces the number of costly distance computations for similar measure. Due to high dimensionality of Motion's features, the Isomap nonlinear dimension reduction is used. An algorithmic framework is used to approximate the optimal mapping function by a Radial Basis Function (RBF) for handling new data. Then a reference index is build based on selecting a small set of representative motion clips in the database. So we can get candidate set by abandoning most unrelated motion clips to reduce the number of costly similarity measure significantly. Experiment results show that our methods are effective for motion data retrieval in large-scale databases.

Paper Details