Voice pathology detection using auto-correlation of different filters bank

Ahmed Al-Nasheri, Zulfiqar Ali, Ghulam Muhammad, Mansour Alsulaiman

Research output: Contribution to conferencePaperpeer-review

20 Citations (Scopus)

Abstract

This paper investigates the contribution of frequency bands for automatic voice pathology detection. First, the input voice signal is passed through a number of time-domain band-pass filters. The center frequencies are spaced on an octave scale. Each filter output is then divided into overlapping frames. Auto-correlation function is applied to each block to find the first largest peak, in areas other than near the dc value, and its corresponding lag. Therefore, each frame is having only these two features (peak value and lag). As classifier, we use Gaussian mixture models (GMM) and support vector machine (SVM), separately. Two well-known available databases, one in English (MEEI) and the other one in German (SVD), are used in the investigation. The results demonstrate that the most significant frequency range to detect voice pathology is between 1500 Hz and 3500 Hz. Using this filter band and with only two features, the accuracy is above 97% in case of the MEEI database.

Original languageEnglish
Pages50-55
Number of pages6
DOIs
Publication statusPublished (in print/issue) - 2 Apr 2015
Event2014 11th IEEE/ACS International Conference on Computer Systems and Applications, AICCSA 2014 - Doha, Qatar
Duration: 10 Nov 201413 Nov 2014

Conference

Conference2014 11th IEEE/ACS International Conference on Computer Systems and Applications, AICCSA 2014
Country/TerritoryQatar
CityDoha
Period10/11/1413/11/14

Keywords

  • Auto-correlation
  • GMM
  • SVD
  • SVM
  • voice pathology detection

Fingerprint

Dive into the research topics of 'Voice pathology detection using auto-correlation of different filters bank'. Together they form a unique fingerprint.

Cite this