Home
People
Projects
Research topics
Publications
Contact
| Professor Unto K. Laine - team leader, D.Sc. |
|
Unto has been a professor in speech technology since 2002. His present research activities are related to
unsupervised pattern learning and modeling (pattern discovery), including study of pattern
theory based on statistical methods, entropy measures and formal languages (context-free
grammars). Applications of pattern theory embrace cognitive architectures, especially language
learning and speech recognition.
Earlier activites: - Development of Finnish text-to-speech system (Synte 2, Synte 3, team member) 1973-1981 - Modeling of vocal tract acoustics 1981-1990 - Frequency warping, warped LP, and FAMlet transfom 1991-1999 - Studies on sounds associated with Aurora Borealis 2000- - Studies of permutation groups applied in pattern recognition 2002- |
| http://www.acoustics.hut.fi/~unski/ |
| Toomas Altosaar, D.Sc. |
|
Toomas has been with TKK's acoustics laboratry since 1985. He received the B. Eng.
(Electrical Engineering) degree from McGill University in 1984 that included graduate
level studies in speech communication. In 1989 he received a Licentiate in Technology
degree from the Helsinki University of Technology (TKK) dealing with Multiresolution
analysis of speech signals. In 2001 a D.Sc. degree dealing with modelling speech
efficiently for database access, also from TKK. Activities in the ACORNS project include modelling the speech recordings created during the project to ensure that EU researchers are able to access computationally subsets of the corpora in a reliable manner. Toomas's interests reside with modelling systems using objects as well as large software systems. Most often he expresses implemented ideas in Lisp. |
| http://www.acoustics.hut.fi/~tal/ |
| Seppo Fagerlund, M.Sc. |
|
Seppo Fagerlund joined the Speech technology team in 2004 to work on on automatic recognition of bird species by their sounds, a topic of which he made his master's thesis in 2004. In addition to continuing the work on bird sound recognition, he is currently also studying permutation transforms and metrics for automatic speech recognition. |
| http://www.acoustics.hut.fi/u/sfagerlu/ |
| Okko Räsänen, M.Sc. |
|
Okko (born in Sept. 1984, Espoo), joined the group in early June 2007 to work in ACORNS project. He received his
Master's degree in Language Technology in November 2007 with a thesis: "Speech Segmentation and Clustering Methods for
a New Speech Recognition Architecture". Currently he is working on his doctoral thesis on new methods for unsupervised language learning. Okko's special interests are computational modeling of language acquisition, statistical learning, and cognitive aspects of language and speech processing in general. |
| http://www.acoustics.hut.fi/~orasanen/ |