Automatic recognition of pitch movements using multilayer perception and time-delay recursive neural network

Sung Sunk Kim, Mark Hasegawa-Johnson, Ken Chen

Research output: Contribution to journalArticle

Abstract

This letter demonstrates hidden Markov model (HMM), multilayer perceptron (MLP), and time-delay recursive neural network (TDRNN) architectures for the purpose of recognizing pitch accents given observation of the F0 and energy trajectories. At an insertion error rate of 25%, the deletion error rates of the MLP, TDRNN, and HMM are 13.2%, 7.9%, and 32.7%, respectively, despite the fact that both MLP and TDRNN have 70% fewer trainable parameters than the HMM. Error analysis suggests that low-pitch accents may require long-term context to correctly recognize, while high-pitch accents may be recognizable based on local pitch contour.

Original languageEnglish (US)
Pages (from-to)645-648
Number of pages4
JournalIEEE Signal Processing Letters
Volume11
Issue number7
DOIs
StatePublished - Jul 1 2004

ASJC Scopus subject areas

  • Signal Processing
  • Electrical and Electronic Engineering
  • Applied Mathematics

Fingerprint Dive into the research topics of 'Automatic recognition of pitch movements using multilayer perception and time-delay recursive neural network'. Together they form a unique fingerprint.

  • Cite this