The search result changed since you submitted your search request. Documents might be displayed in a different sort order.
  • search hit 20 of 94
Back to Result List

Prosodic addressee-detection : ensuring privacy in always-on spoken dialog systems

  • We analyze the addressee detection task for complexity-identical dialog for both human conversation and device-directed speech. Our recurrent neural model performs at least as good as humans, who have problems with this task, even native speakers, who profit from the relevant linguistic skills. We perform ablation experiments on the features used by our model and show that fundamental frequency variation is the single most relevant feature class. Therefore, we conclude that future systems can detect whether they are addressed based only on speech prosody which does not (or only to a very limited extent) reveal the content of conversations not intended for the system.

Export metadata

Additional Services

Share in Twitter Search Google Scholar Statistics
Metadaten
Author:Timo BaumannORCiDGND, Ingo Siegert
DOI:https://doi.org/10.1145/3404983.3410021
ISBN:9781450375405
Parent Title (English):Tagungsband Mensch und Computer (MuC'20): 06.09.2020 - 09.09.2020, Magdeburg
Publisher:Association for Computing Machinery
Place of publication:New York, NY, United States
Editor:Florian Alt, Stefan Schneegass, Eva Hornecker
Other Person(s):Bernhard Preim
Document Type:conference proceeding (article)
Language:English
Year of first Publication:2020
Release Date:2022/04/09
Tag:addressee detection; complexity-identical human-computer interaction; computational paralinguistics; fundamental frequency variation; recurrent neural network
First Page:195
Last Page:198
Institutes:Fakultät Informatik und Mathematik
Begutachtungsstatus:peer-reviewed
Publication:Externe Publikationen
research focus:Information und Kommunikation