Predicting interviewee attitude and body language from speech descriptors
This work addresses the problem of automated assessment of interviewee profiles for HR or psychology applications, but it is incremental as it builds on existing methods for analyzing speech cues.
The study tackled predicting interviewee attitudes and body language from acoustic speech descriptors, finding that topicality caused significant deviations in acoustic parameters, but prediction of personal perceptions was relatively independent of topicality due to redundancy in acoustic attributes.
This present research investigated the relationship between personal impressions and the acoustic nonverbal communication conveyed by employees being interviewed. First, we investigated the extent to which different conversation topics addressed during the interview induced changes in the interviewees' acoustic parameters. Next, we attempted to predict the observed and self-assessed attitudes and body language of the interviewees based on the acoustic data. The results showed that topicality caused significant deviations in the acoustic parameters statistics, but the ability to predict the personal perceptions of the interviewees based on their acoustic non-verbal communication was relatively independent of topicality, due to the natural redundancy inherent in acoustic attributes. Our findings suggest that joint modeling of speech and visual cues may improve the assessment of interviewee profiles.