In exemplar-based approaches for human pose estimation, it is common to extract multiple features to better describe the visual input data. However, simply concatenating multiview features into a long vector has two shortcomings: (1) it suffers from "curse of dimensionality"; (2) it is not physically meaningful and may be incapable of fully exploiting the complementary properties of multi-view features. To address such problems, in this paper we present a dimension reduction method based on supervised spectral embedding, followed by an ensemble of nearest neighbor regressions in multi-view feature space, to infer 3D human poses from monocular videos. The experiments on HumanEva dataset show the effectiveness of the proposed method.
CITATION STYLE
Guo, Y., Chen, Z., & Yu, J. (2015). Supervised spectral embedding for human pose estimation. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9242, pp. 100–109). Springer Verlag. https://doi.org/10.1007/978-3-319-23989-7_11
Mendeley helps you to discover research relevant for your work.