Improving SVF with DISTBIC for Phoneme Segmentation

Abstract : In this paper we examine an application for phoneme segmentation of DISTBIC, a two-pass, textindependent method traditionally used for speaker segmentation. The novelty of this paper is its experimentation with use of the spectral variation function (SVF), a simple non-parametric method for phone segmentation, as a replacement for the distance measure of the first pass of DISTBIC. In doing so we aim to produce a computationally efficient method for text-independent phoneme segmentation that provides good performance. Experiments are carried out on the TIMIT database. We give a performance comparison between the SVF as previously used for segmentation, our DISTBIC-SVF algorithm, and another state-of-the-art algorithm.
Complete list of metadatas

Cited literature [10 references]  Display  Hide  Download

https://hal.inria.fr/hal-00647985
Contributor : Khalid Daoudi <>
Submitted on : Sunday, December 4, 2011 - 3:49:21 PM
Last modification on : Wednesday, November 14, 2018 - 1:54:02 PM
Long-term archiving on: Monday, March 5, 2012 - 2:20:39 AM

File

SPECOM-2011.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00647985, version 1

Collections

Citation

Joshua Winebarger, Khalid Daoudi, Hussein Yahia. Improving SVF with DISTBIC for Phoneme Segmentation. International Conference Speech and Computer (SPECOM), Sep 2011, Kazan, Russia. ⟨hal-00647985⟩

Share

Metrics

Record views

411

Files downloads

162