Research Student Project
Supervisor name & contact details:
Name: Dr. Brian Vaughan
School : Media
Email: brian.vaughan@dit.ie
Weblink (if available): istory.ie
Research Centre Name and Website (if
applicable)
Digital Media Centre
Please indicate if the intention is to transfer
from the Masters programme to the PhD
programme (if applicable)
No
Please indicate if the project is suitable for a
self-funded student
No (funding in place for the student)
Funding Agency
Dublin Institute of Technology
Scholarship Details
€16k per year with €2k travel and costs for four
years. Fees are paid.
Social signal processing
Subject Area
Title of the Project
Depression Diagnosis and Treatment Through
Speech (DDaTTS)
Project Description (max 300 words)
The goal of this research is to examine how depression affects a person’s speech in order to
develop tools and methods for the detection and treatment of depressive illness. The research
will build upon existing research on depressed speech and will examine important sociocommunicative non-verbal speech characteristics, such as prosodic accommodation, that can be
used to measure rapport and involvement between speakers. This has the potential to provide
advanced insights into the depressive state of a patient, including a determination of depression
severity. The outcomes of the research will further be used to develop mobile applications to
augment current clinical practices, and to aid and monitor depressed patients in a non-clinical
setting.
Depression is a relatively common phenomenon, affecting two in ten people in Ireland during
their lifetime. Research by the World Health Organisation states that 1 in 15 people across the
WHO European Region currently suffer from depression, which is approximately 49 Million
people. It is an affective disorder, characterised by a number of different factors that may be
present in a number of different combinations and severities.
People engaged in conversation tend to adapt their communicative behaviour to that of their
interlocutor, with numerous terms being used to describe this phenomenon, particularly with
regards to speech: accommodation, synchrony, mimicry, convergence, alignment and
entrainment. Vocal accommodation occurs between speakers when changes in their prosodic
parameters move in synchronous alignment or when they converge towards a common point; it
is a particularly important aspect of social interaction as it facilitates comprehension and
understanding between interlocutors. In addition, accommodation participates in increasing the
social success of the interaction in terms of rapport (i.e. harmonious relation and mutual
attention) and affiliation. Numerous studies have found that depression induces measurable,
manifest, changes in a person’s speech that can be indicative of depression severity or even
suicidal ideation.
Please indicate the student requirements for this project
Min 2.1 BSc in Computer Science or related technical domain (electrical engineering, physics etc)
Ideally: MSc in Computer Science or related technical domain (electrical engineering, physics etc)
Must have experience with signal processing (DSP) visual or auditory.
Experience using Matlab and/or Octave desirable
Familiar with one or more of the following programming languages: Python, java, C#, C++, C
Candidates from a non-scientific/technical background (BA and MA) may be considered based on
RPL, experience, and skillset.
Deadline to submit applications (only for
funded projects)
1st August 2015