Learning Kernel-based HMMs for dynamic sequence synthesis

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

4 Scopus citations

Abstract

In this paper we present an approach that synthesizes a dynamic sequence from another related sequence, and apply it to a virtual conductor: to synthesize linked figure animation from an input music track. We propose that the mapping between two dynamic sequences can be modeled with a Kernel-based Hidden Markov model, or KHMM. A KHMM is an HMM for which the kernel-based functions are used to model the state observation density of the joint input and output distribution. Specifically, the state observation density is estimated by employing a likelihood-weighted sampling scheme. Our KHMM model is ideal for dynamic sequence synthesis because the global dynamics are learned by the HMM, and subtle details in the dynamic mapping are kept in the kernel-based state density. We demonstrate our virtual conductor by synthesizing extensive animation sequences from input music sequences with different styles and beat patterns.

Original languageEnglish
Title of host publicationProceedings - 10th Pacific Conference on Computer Graphics and Applications, PG 2002
EditorsShi-Min Hu, Heung-Yeung Shum, Sabine Coquillart
PublisherIEEE Computer Society
Pages87-95
Number of pages9
ISBN (Electronic)0769517846
DOIs
StatePublished - 2002
Event10th Pacific Conference on Computer Graphics and Applications, PG 2002 - Beijing, China
Duration: 9 Oct 200211 Oct 2002

Publication series

NameProceedings - Pacific Conference on Computer Graphics and Applications
Volume2002-January
ISSN (Print)1550-4085

Conference

Conference10th Pacific Conference on Computer Graphics and Applications, PG 2002
Country/TerritoryChina
CityBeijing
Period9/10/0211/10/02

Keywords

  • Animation
  • Asia
  • Conductors
  • Hidden Markov models
  • Humans
  • Kernel
  • Motion analysis
  • Signal synthesis
  • Speech synthesis
  • Video sharing

Fingerprint

Dive into the research topics of 'Learning Kernel-based HMMs for dynamic sequence synthesis'. Together they form a unique fingerprint.

Cite this