Defining Ally’s Interactions

A Deep Learning framework to create a personalised interaction between users and a medical pod

Master Thesis (2018)
Author(s)

S.R. Rangarajan (TU Delft - Industrial Design Engineering)

Contributor(s)

Maaike S. Kleinsmann – Mentor

A.Q. Beekman – Coach

Faculty
Industrial Design Engineering
Copyright
Campus only
More Info
expand_more
Publication Year
2018
Language
English
Copyright
Campus only
Graduation Date
29-03-2018
Awarding Institution
Delft University of Technology
Programme
Design for Interaction
Faculty
Industrial Design Engineering
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Ally, is an intelligent, voice activated medical device concept that users talk to on a routine basis to log details regarding their health and well-being. The goal of this thesis is to create a personalised interaction between users and Ally. The user group is segregated into three generations; Baby Boomers, Generation X and Millennials. A digital prototype of Ally is used to understand how different generations interact with Ally. A questionnaire to identify the voice characteristics each generation prefers in a medical device, is created. The results from the user test and questionnaire are used to design a Deep Learning framework to generate a WaveNet TTS voice. This framework is a foundation for a personalised interaction between Ally and users, based on the generation they belong to. By creating a framework to cater to specific generations, this model sets the ground rules for personalisation.

Files

License info not available

Download not available

License info not available

Download not available

License info not available

Download not available

Framework.jpg
(jpg | 0.135 Mb)
License info not available

Download not available

License info not available

Download not available