Analysis of emotional speech at subsegmental level

Gangamohan, P.; Kadiri, Sudarsana Reddy; Yegnanarayana, B.

doi:10.21437/interspeech.2013-20

Cited by 21 publications

(5 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These features were shown to be useful for discriminating phonation types in speech and singing [13], [41], [42]. SoE was shown to be proportional to the rate of glottal closure, the EoE feature was shown to capture the vocal effort, and the loudness measure was shown to capture the abruptness of the glottal closure [43], [44]. The energy of the ZFF signal at glottal closure is also used as a feature, it was shown to capture low frequency energy [13].…”

Section: ) Zero Frequency Filtering (Zff)mentioning

confidence: 99%

Classification of Phonation Modes in Classical Singing Using Modulation Power Spectral Features

et al. 2023

Self Cite

View full text Add to dashboard Cite

In singing, the perceptual term "voice quality" is used to describe expressed emotions and singing styles. In voice physiology research, specific voice qualities are discussed by the term phonation modes and are related directly to the voicing produced by the vocal folds. The control and awareness of phonation modes is vital for professional singers to maintain a healthy voice. Most studies on phonation modes have investigated speech and have used glottal inverse filtering to compute features from an estimated excitation signal. The performance of this method is reported to decrease at high pitches, which limits its usability for the singing voice. To overcome this, this study proposes to use features derived from the modulation power spectrum for phonation mode classification in the singing voice. The exploration of the modulation power spectrum is motivated by the fact that, in singing, temporal modulations (known as vocal vibrato) and spectral modulations hold information of the vocal fold tension. Since there exists no large publicly available dataset of phonation modes in singing, we created a new dataset consisting of six female and four male classical singers, who sang five vowels at different pitches in three phonation modes (breathy, modal, and pressed). Experimental results with a support vector machine classifier reveal that the proposed features show better classification performance compared to state-of-the-art reference features. The performance for the current dataset is at least 10% higher compared to the performance of the reference features (such as glottal source features and MFCCs) in the case of target labels and around 6% higher in the case of perceptually assessed labels.INDEX TERMS Modulation power spectrum, phonation modes, singing voice analysis, voice qualities.

show abstract

Section: ) Zero Frequency Filtering (Zff)mentioning

confidence: 99%

Classification of Phonation Modes in Classical Singing Using Modulation Power Spectral Features

et al. 2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…This feature was used in the analysis and classification of vocal emotions in [58,59], where the feature was shown to be large for emotions of low arousal (such as sadness), and small for emotions of high arousal (such as anger and happiness). Therefore, this feature reflects changes in the relative duration of the glottal closed phase in a similar manner to CQ and NAQ [16,60].…”

Section: Features Derived Using Zffmentioning

confidence: 99%

“…where 2K+1 corresponds to the number of samples in the 1-ms window. This feature was shown to reflect the changes in vocal effort [58,59]. The experiments in [58,59] indicated that EoE was generally large for emotions of high arousal and small for emotions of low arousal.…”

Section: Features Derived Using Zffmentioning

confidence: 99%

Analysis and classification of phonation types in speech and singing voice

Kadiri

Alku

Yegnanarayana

2020

Speech Communication

Self Cite

View full text Add to dashboard Cite

“…The fundamental frequency (F0 = 1/T0) is a popular acoustic feature that can be estimated from speech using pitch analysis methods, but other source parameters that are correlated with emotions can also be measured from speech, e.g. the strength of excitation and energy of excitation [2,3], aperiodicity features (jitter, shimmer, etc.) [4,5] and energy ratio between different frequency bands [6].…”

Section: Introductionmentioning

confidence: 99%

Voice source correlates of acted male speech emotions

Cabral,

Govind

2023

SMM23, Workshop on Speech, Music and Mind 2023

View full text Add to dashboard Cite

The glottal source signal conveys very important information about the speech emotions. For example, source parameters have been successfully used in applications of recognition, transformation and synthesis of speech emotions. Often the source parameters are extracted from the speech signal only, without estimating the glottal source component, mainly because of the difficulty to separate this component from speech. This paper aims to improve the knowledge about the acoustic correlates of emotions by using accurate measurements of shape parameters of the glottal pulse. To achieve this, the most relevant glottal pulse instants are estimated using a semi-automatic method based on the analysis of both the electroglottographic and speech signals. Additional source parameters, the strength of excitation and harmonic-to-noise ratio, are also estimated from the speech signal, for comparison with the glottal parameters in our experiment. The average results of analysis of the source parameters show that they generally have a strong correlation with the set of emotions using neutral as reference: angry, disgusted, happy, sad, scared, and surprised. A possible application of this study is to derive parameter scale factors from the average measurements of the glottal parameters for speech emotion transformation.

show abstract

Analysis of emotional speech at subsegmental level

Cited by 21 publications

References 12 publications

Classification of Phonation Modes in Classical Singing Using Modulation Power Spectral Features

Classification of Phonation Modes in Classical Singing Using Modulation Power Spectral Features

Analysis and classification of phonation types in speech and singing voice

Voice source correlates of acted male speech emotions

Contact Info

Product

Resources

About