Multimodal analysis of vocal emotion and prosody in subclinical depressed individuals
摘要
This study aimed to investigate the characteristics of vocal emotion and prosody in individuals with subclinical depression. A total of 70 college students were assigned to either the subclinical group or the control group and participated in a videotaped face-to-face interview. The Beck Depression Inventory (BDI-II-C) and the Patient Health Questionnaire (PHQ-9) were utilized to screen the participants. The interview videos were segmented and annotated tier-by-tier through Elan software. The tiers in this study included vocal emotions and prosody. The results demonstrated that compared with the control group, the subclinical group exhibited more diverse changes in prosody, and faster changes. The most prominent co-occurrence between vocal emotion and prosody was “pause” and “hesitation” in the subclinical group. “Raise the tone suddenly” was high co-occurrence with “excitement” and “delight”, and “pause” was high co-occurrence with “delight”, and “raise the tone” was high co-occurrence with “excitement” and “delight” in the control group, these high co-occurrence networks were connected with the main node of “delight”. The features of vocal emotion and prosody could be the potential indicators of depression.