From Joyous to Clinically Depressed: Mood Detection Using Spontaneous Speech

Sharifa Alghowinem, Roland Goecke, Michael Wagner, Julien Epps, Michael Breakspear, Gordon Parker

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    34 Citations (Scopus)

    Abstract

    Depression and other mood disorders are common and disabling disorders. We present work towards an objective diagnostic aid supporting clinicians using affective sensing technology with a focus on acoustic and statistical features from spontaneous speech. This work investigates differences in expressing positive and negative emotions in depressed and healthy control subjects as well as whether initial gender classification increases the recognition rate. To this end, spontaneous speech from interviews of 30 subjects of each depressed and controls was analysed, with a focus on questions eliciting positive and negative emotions. Using HMMs with GMMs for classification with 30-fold cross-validation, we found that MFCC, energy and intensity features gave highest recognition rates when female and male subjects were analysed together. When the dataset was first split by gender, log energy and shimmer features, respectively, were found to give the highest recognition rates in females, while it was loudness for males. Overall, correct recognition rates from acoustic features for depressed female subjects were higher than for male subjects. Using temporal features, we found that the response time and average syllable duration were longer in depressed subjects, while the interaction involvement and articulation rate wesre higher in control subjects.
    Original languageEnglish
    Title of host publicationProceedings of the Twenty-Fifth International Florida Artificial Intelligence Research Society Conference
    EditorsG.M Youngblood, P.M McCarthy
    Place of PublicationFlorida
    PublisherAssociation for the Advancement of Artificial Intelligence (AAAI)
    Pages141-146
    Number of pages6
    ISBN (Electronic)9781577355595
    ISBN (Print)9781577355588
    Publication statusPublished - 2012
    EventInternational Florida Artificial Intelligence Research Society Conference (FLAIRS 2012) - Marco Island, Marco Island, United States
    Duration: 23 May 201225 May 2012
    http://projects.ict.usc.edu/flairs-25/

    Conference

    ConferenceInternational Florida Artificial Intelligence Research Society Conference (FLAIRS 2012)
    Abbreviated titleFLAIRS 2012
    CountryUnited States
    CityMarco Island
    Period23/05/1225/05/12
    Internet address

    Fingerprint

    Acoustics

    Cite this

    Alghowinem, S., Goecke, R., Wagner, M., Epps, J., Breakspear, M., & Parker, G. (2012). From Joyous to Clinically Depressed: Mood Detection Using Spontaneous Speech. In G. M. Youngblood, & P. M. McCarthy (Eds.), Proceedings of the Twenty-Fifth International Florida Artificial Intelligence Research Society Conference (pp. 141-146). Florida: Association for the Advancement of Artificial Intelligence (AAAI).
    Alghowinem, Sharifa ; Goecke, Roland ; Wagner, Michael ; Epps, Julien ; Breakspear, Michael ; Parker, Gordon. / From Joyous to Clinically Depressed: Mood Detection Using Spontaneous Speech. Proceedings of the Twenty-Fifth International Florida Artificial Intelligence Research Society Conference. editor / G.M Youngblood ; P.M McCarthy. Florida : Association for the Advancement of Artificial Intelligence (AAAI), 2012. pp. 141-146
    @inproceedings{2d310a5d7eec4e1c8dd734022012f53d,
    title = "From Joyous to Clinically Depressed: Mood Detection Using Spontaneous Speech",
    abstract = "Depression and other mood disorders are common and disabling disorders. We present work towards an objective diagnostic aid supporting clinicians using affective sensing technology with a focus on acoustic and statistical features from spontaneous speech. This work investigates differences in expressing positive and negative emotions in depressed and healthy control subjects as well as whether initial gender classification increases the recognition rate. To this end, spontaneous speech from interviews of 30 subjects of each depressed and controls was analysed, with a focus on questions eliciting positive and negative emotions. Using HMMs with GMMs for classification with 30-fold cross-validation, we found that MFCC, energy and intensity features gave highest recognition rates when female and male subjects were analysed together. When the dataset was first split by gender, log energy and shimmer features, respectively, were found to give the highest recognition rates in females, while it was loudness for males. Overall, correct recognition rates from acoustic features for depressed female subjects were higher than for male subjects. Using temporal features, we found that the response time and average syllable duration were longer in depressed subjects, while the interaction involvement and articulation rate wesre higher in control subjects.",
    keywords = "Depression detection, Speaker Charracterisation",
    author = "Sharifa Alghowinem and Roland Goecke and Michael Wagner and Julien Epps and Michael Breakspear and Gordon Parker",
    year = "2012",
    language = "English",
    isbn = "9781577355588",
    pages = "141--146",
    editor = "G.M Youngblood and P.M McCarthy",
    booktitle = "Proceedings of the Twenty-Fifth International Florida Artificial Intelligence Research Society Conference",
    publisher = "Association for the Advancement of Artificial Intelligence (AAAI)",

    }

    Alghowinem, S, Goecke, R, Wagner, M, Epps, J, Breakspear, M & Parker, G 2012, From Joyous to Clinically Depressed: Mood Detection Using Spontaneous Speech. in GM Youngblood & PM McCarthy (eds), Proceedings of the Twenty-Fifth International Florida Artificial Intelligence Research Society Conference. Association for the Advancement of Artificial Intelligence (AAAI), Florida, pp. 141-146, International Florida Artificial Intelligence Research Society Conference (FLAIRS 2012), Marco Island, United States, 23/05/12.

    From Joyous to Clinically Depressed: Mood Detection Using Spontaneous Speech. / Alghowinem, Sharifa; Goecke, Roland; Wagner, Michael; Epps, Julien; Breakspear, Michael; Parker, Gordon.

    Proceedings of the Twenty-Fifth International Florida Artificial Intelligence Research Society Conference. ed. / G.M Youngblood; P.M McCarthy. Florida : Association for the Advancement of Artificial Intelligence (AAAI), 2012. p. 141-146.

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    TY - GEN

    T1 - From Joyous to Clinically Depressed: Mood Detection Using Spontaneous Speech

    AU - Alghowinem, Sharifa

    AU - Goecke, Roland

    AU - Wagner, Michael

    AU - Epps, Julien

    AU - Breakspear, Michael

    AU - Parker, Gordon

    PY - 2012

    Y1 - 2012

    N2 - Depression and other mood disorders are common and disabling disorders. We present work towards an objective diagnostic aid supporting clinicians using affective sensing technology with a focus on acoustic and statistical features from spontaneous speech. This work investigates differences in expressing positive and negative emotions in depressed and healthy control subjects as well as whether initial gender classification increases the recognition rate. To this end, spontaneous speech from interviews of 30 subjects of each depressed and controls was analysed, with a focus on questions eliciting positive and negative emotions. Using HMMs with GMMs for classification with 30-fold cross-validation, we found that MFCC, energy and intensity features gave highest recognition rates when female and male subjects were analysed together. When the dataset was first split by gender, log energy and shimmer features, respectively, were found to give the highest recognition rates in females, while it was loudness for males. Overall, correct recognition rates from acoustic features for depressed female subjects were higher than for male subjects. Using temporal features, we found that the response time and average syllable duration were longer in depressed subjects, while the interaction involvement and articulation rate wesre higher in control subjects.

    AB - Depression and other mood disorders are common and disabling disorders. We present work towards an objective diagnostic aid supporting clinicians using affective sensing technology with a focus on acoustic and statistical features from spontaneous speech. This work investigates differences in expressing positive and negative emotions in depressed and healthy control subjects as well as whether initial gender classification increases the recognition rate. To this end, spontaneous speech from interviews of 30 subjects of each depressed and controls was analysed, with a focus on questions eliciting positive and negative emotions. Using HMMs with GMMs for classification with 30-fold cross-validation, we found that MFCC, energy and intensity features gave highest recognition rates when female and male subjects were analysed together. When the dataset was first split by gender, log energy and shimmer features, respectively, were found to give the highest recognition rates in females, while it was loudness for males. Overall, correct recognition rates from acoustic features for depressed female subjects were higher than for male subjects. Using temporal features, we found that the response time and average syllable duration were longer in depressed subjects, while the interaction involvement and articulation rate wesre higher in control subjects.

    KW - Depression detection

    KW - Speaker Charracterisation

    M3 - Conference contribution

    SN - 9781577355588

    SP - 141

    EP - 146

    BT - Proceedings of the Twenty-Fifth International Florida Artificial Intelligence Research Society Conference

    A2 - Youngblood, G.M

    A2 - McCarthy, P.M

    PB - Association for the Advancement of Artificial Intelligence (AAAI)

    CY - Florida

    ER -

    Alghowinem S, Goecke R, Wagner M, Epps J, Breakspear M, Parker G. From Joyous to Clinically Depressed: Mood Detection Using Spontaneous Speech. In Youngblood GM, McCarthy PM, editors, Proceedings of the Twenty-Fifth International Florida Artificial Intelligence Research Society Conference. Florida: Association for the Advancement of Artificial Intelligence (AAAI). 2012. p. 141-146