Skip to main navigation Skip to search Skip to main content

Supervised spectral embedding for human pose estimation

  • Yukun Guo
  • , Zhonggui Chen
  • , Jun Yu*
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

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.

Original languageEnglish
Title of host publicationIntelligence Science and Big Data Engineering
Subtitle of host publicationImage and Video Data Engineering - 5th International Conference, IScIDE 2015, Revised Selected Papers
EditorsXiaofei He, Zhi-Hua Zhou, Xinbo Gao, Zhi-Yong Liu, Yanning Zhang, Baochuan Fu, Fuyuan Hu, Zhancheng Zhang
PublisherSpringer Verlag
Pages100-109
Number of pages10
ISBN (Print)9783319239873
DOIs
StatePublished - 2015
Externally publishedYes
Event5th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2015 - Suzhou, China
Duration: 14 Jun 201516 Jun 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9242
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference5th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2015
Country/TerritoryChina
CitySuzhou
Period14/06/1516/06/15

Keywords

  • Human pose estimation
  • K-NN regression
  • Spectral embedding

Fingerprint

Dive into the research topics of 'Supervised spectral embedding for human pose estimation'. Together they form a unique fingerprint.

Cite this