Skip to Main content Skip to Navigation
Preprints, Working Papers, ...

Multi-view pose estimation with mixtures-of-parts and adaptive viewpoint selection

Abstract : We propose a new method for human pose estimation which leverages information from multiple views to impose a strong prior on articulated pose. The novelty of the method concerns the types of coherence modelled. Consistency is maximised over the different views through different terms modelling classical geometric information (coherence of the resulting poses) as well as appearance information which is modelled as latent variables in the global energy function. Moreover, adequacy of each view is assessed and their contributions are adjusted accordingly. Experiments on the HumanEva and UMPM datasets show that the proposed method significantly decreases the estimation error compared to single-view results.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01593547
Contributor : Christian Wolf <>
Submitted on : Tuesday, September 26, 2017 - 2:12:16 PM
Last modification on : Thursday, November 21, 2019 - 1:43:48 AM

Links full text

Identifiers

  • HAL Id : hal-01593547, version 1
  • ARXIV : 1709.08527

Citation

Emre Dogan, Gonen Eren, Christian Wolf, Eric Lombardi, Atilla Baskurt. Multi-view pose estimation with mixtures-of-parts and adaptive viewpoint selection. 2017. ⟨hal-01593547⟩

Share

Metrics

Record views

197