Toward detecting and classification non-verbal events and biosignals in hearables

Auteurs-es

  • Malahat Mehrban Université du Québec (ÉTS)<br />
  • Jérémie Voix Université du Québec (ÉTS)<br />
  • Rachel Bouserhal Université du Québec (ÉTS)<br />

Résumé

In today’s hectic world, people are always searching for more ways of keeping fit and healthy. For this reason, using audio wearable devices that record physiological events from human body is becoming increasingly popular. Within the stable position of the ear canal, microphones could record heart and respiratory sounds. Due to the heavy computation of conventional signal processing methods which are based on statistical algorithms, fast and reliable alternatives are required to be used in tiny hearables that have limited power sources and processors. Since machine learning based methods are popular for their rapid computational and comprehensive learning capacity, could be an appropriate alternative to conventional methods. The aim of this project is to work on the data collected from in-ear microphones. Upon successful completion of this experiment, it is expected to detect and classify different physiological events to monitor the wearers' health and emotions.

Fichiers supplémentaires

Publié-e

2022-08-19

Comment citer

1.
Mehrban M, Voix J, Bouserhal R. Toward detecting and classification non-verbal events and biosignals in hearables. Canadian Acoustics [Internet]. 19 août 2022 [cité 3 nov. 2024];50(3):76-7. Disponible à: https://jcaa.caa-aca.ca/index.php/jcaa/article/view/3886

Numéro

Rubrique

Actes du congrès de la Semaine canadienne d'acoustique