Supervised spectral embedding for human pose estimation

Abstract

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.