Laboratory of Acoustics and Audio Signal Processing
Helsinki University of Technology, FI
Experience and Competence
In 1980, the first professorship was established in the Laboratory of Acoustics
and Audio Signal Processing at the Helsinki University of Technology (HUT).
Currently, there are three professors in the unit and the total number
of personnel equals 27. The major themes in the laboratory's research activities
are virtual acoustics, 3-D sound technology, musical acoustics and speech
processing, especially modelling of voice production, speech recognition
and development of speech databases.
Modelling of speech production has been a topic of research of HUT
for more than ten years. During these years the laboratory has gained expertise
especially in the estimation of the glottal flow with inverse filtering
techniques and in the development of parameterisation methods of voice
production. In this area, the Laboratory of Acoustics and Audio Signal
Processing has been in active collaboration with Finnish medical universities.
In 1997, the Laboratory of Acoustics and Audio Signal Processing started
to collaborate also with scientists of cognitive brain research. In this
area, the collaboration has been intensive especially with the Department
of Psychology at the University of Helsinki, but there has also been co-operation
with other European universities. Therefore, the Laboratory of Acoustics
and Audio Signal Processing at HUT is well-suited to participate in the
proposed project, because, firstly, the laboratory owns state-of-the-art
technical expertise in the topic of the proposed task and the scientific
output on modelling of human speech production has been rich during the
past decade. Secondly, the laboratory has a long experience in performing
interdisciplinary research with scientists representing such areas as phoniatrics
and cognitive brain research. This kind of experience is an important pre-requisite
to succeed in the proposed research project.
Role in the Project
HUT will lead theme 3, speech production modelling.
HUT is already linked to the other participants of the proposed project
through one researcher (Kalle Palomäki, MSc), who works at the University
of Sheffield in the SPHEAR program.
Principal research personnel,
expertise and involvement
Paavo Alku, Professor, D.Sc (Tech)., Expertise in inverse filtering and
parameterisation of voice production, involvement 10 man-months.
Matti Karjalainen, Professor, D.Sc (Tech)., Expertise in 3D signal processing
and binaural modelling . involvement 5 man-months.
Alku, P., Tiitinen, H., Näätänen, R. "A method for generating
natural-sounding speech stimuli for cognitive brain research", Clinical
Neurophysiology, Vol. 110, pp. 1329-1333, 1999.
Alku, P., Sivonen, P., Palomäki, K., Tiitinen, H. "The periodic structure
of vowel sounds is reflected in human electromagnetic brain responses",
Neuroscience Letters, Vol. 298, No. 1, pp. 25-28, 2001.