As computer-generated voice synthesis has become a significant part of communications between computers and people, there is a need to understand the role of paralinguistic attributes of the voice, such as age, personality, and gender. In many cases, the synthesized voice is produced by concatenating segments of recorded human speech, which can be experienced as a lifeless voice that lacks free expression and fluidness. Technology companies have been developing their own unique synthesized voice identities without paying attention to the stereotypical traits being heard. This study evaluated the responses of 18 listeners who were asked to consider the paralinguistic traits of age, gender, and human likeness from 13 voices in IBM’s Watson corpus. The results of this study were similar to a previous study, with no voice achieving complete human likeness, no voice being perceived within a single age frequency band, and none tied solidly to their given binary gender.
Authors:
Baird, Alice; Jørgensen, Stina Hasse; Parada-Cabaleiro, Emilia; Cummins, Nicholas; Hantke, Simone; Schuller, Björn
Affiliations:
ZD.B Chair of Embedded Intelligence for Health Care and Wellbeing, University of Augsburg, Germany; Department of Arts & Cultural Studies, University of Copenhagen, Denmark; MISP Group, MKK Technische Universität München, Germany; GLAM, Imperial College London, UK(See document for exact affiliation information.)
JAES Volume 66 Issue 4 pp. 277-285; April 2018
Publication Date:
April 29, 2018
Click to purchase paper as a non-member or you can login as an AES member to see more options.
No AES members have commented on this paper yet.
To be notified of new comments on this paper you can subscribe to this RSS feed. Forum users should login to see additional options.
If you are not yet an AES member and have something important to say about this paper then we urge you to join the AES today and make your voice heard. You can join online today by clicking here.