The feasibility of automatically detecting cardiovascular reactivity from speech was investigated. There are studies that have shown success in detecting heart rate in the speech signal before but cardiovascular reactivity has not been looked at as well. Gender-specific, speaker-independent Gaussian mixture models were trained on speech during high and low cardiovascular reactivity and classification implemented using a cosine distance scoring (ivector) approach. Using five distinct criteria to determine whether classification was meaningful, we found clear indication that cardiovascular reactivity affects the voice in a manner that makes it automatically detectable in speech. As such it may become a powerful new information source for estimating various physiological conditions from speech.
|Number of pages||5|
|Journal||Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH|
|Publication status||Published - 2015|
|Event||16th Annual Conference of the International Speech Communication Association, INTERSPEECH 2015 - Dresden, Germany|
Duration: 6 Sept 2015 → 10 Sept 2015
Bibliographical notePublisher Copyright:
Copyright © 2015 ISCA.
- Biometric data
- Cardiovascular monitoring
- Cognitive load
- Emotion recognition
- Speech analysis