Acoustical correlates of affective prosody

Area of Science:

• Speech analysis
• Acoustic phonetics
• Emotion recognition

Background:

• Understanding the acoustic correlates of human emotions is crucial for fields like psychology and human-computer interaction.
• Previous research has explored vocal biomarkers for emotions, but comprehensive acoustic profiling remains an active area of investigation.

Purpose of the Study:

• To identify specific acoustic parameters that characterize six distinct emotional expressions in speech.
• To determine acoustic indicators that differentiate aversive from hedonistic emotions.

Main Methods:

• Acoustic analysis of the spoken word "Anna" by 23 subjects expressing six emotions.
• Extraction and statistical analysis of 94 acoustic parameters, reduced to 15 key features.
• Correlation analysis to identify significant acoustic profiles for each emotion.

Main Results:

• Each of the six emotions exhibited a unique and statistically significant acoustic signature.
• The ratio of peak frequency to fundamental frequency was the strongest indicator of aversive emotions.
• Other parameters like peak frequency and noise percentage also differentiated emotion types.

Conclusions:

• Acoustic analysis can reliably distinguish between different emotional states in speech.
• Specific vocal features serve as robust biomarkers for emotional expression, particularly for differentiating negative from positive affect.
• The ratio of peak to fundamental frequency offers a loudness-independent measure for aversive emotions.