An Investigation of Depressed Speech Detection: Features and Normalization

Nicholas Cummins, Julien Epps, Michael Breakspear, Roland Goecke

Research output: A Conference proceeding or a Chapter in BookConference contributionpeer-review

140 Citations (Scopus)
949 Downloads (Pure)

Abstract

In recent years, the problem of automatic detection of mental illness from the speech signal has gained some initial interest, however questions remaining include how speech segments should be selected, what features provide good discrimination, and what benefits feature normalization might bring given the speakerspecific nature of mental disorders. In this paper, these questions are addressed empirically using classifier configurations employed in emotion recognition from speech, evaluated on a 47-speaker depressed/neutral read sentence speech database. Results demonstrate that (1) detailed spectral features are well suited to the task, (2) speaker normalization provides benefits mainly for less detailed features, and (3) dynamic information appears to provide little benefit. Classification accuracy using a combination of MFCC and formant based features approached 80% for this database.
Original languageEnglish
Title of host publicationINTERSPEECH 2011 12th Annual Conference of the International Speech Comm. Assoc.
EditorsPiero Cosi, Renato De Mori, Giuseppe Di Fabbrizio, Roberto Pieraccini
Place of PublicationFlorence, Italy
PublisherInternational Speech Communication Association
Pages2997-3000
Number of pages4
ISBN (Print)9781618392701
Publication statusPublished - 27 Aug 2011
EventINTERSPEECH 2011 12th Annual Conference of the International Speech Communication Association - Florence, Florence, Italy
Duration: 27 Aug 201131 Aug 2011

Conference

ConferenceINTERSPEECH 2011 12th Annual Conference of the International Speech Communication Association
Country/TerritoryItaly
CityFlorence
Period27/08/1131/08/11

Fingerprint

Dive into the research topics of 'An Investigation of Depressed Speech Detection: Features and Normalization'. Together they form a unique fingerprint.

Cite this