MIKKO KURIMO, D.Sc.(Tech.)
My research interests include machine learning, speech recognition, information retrieval and natural language processing. In 1998 - 2000 I worked as a post doc at IDIAP in Switzerland doing research on large vocabulary continuous speech recognition and indexing broadcast news by speech recognition. There I participated in two long-term European projects: THISL (Thematic Indexing of Spoken Language) and ASSAVID (Automatic Segmentation and Semantic Annotation of Sports Videos). Then I returned to the Neural Networks Research Centre in order to improve our continuous speech recognition and use it for speech indexing. My research group has developed a fully home-grown unlimited vocabulary speech recognition system for Finnish and tested it also for Estonian and English. Especially, the statistical (adaptive) morpheme-based language modeling for the agglutinative languages, like Finnish, Estonian and Turkish, has been our major achievement. I am also involved in a few industrial projects as well as the projects of Finnish Technology Programs USIX(User-Oriented Information Technology) and FENIX (Interactive Computing) and the development of the new Finnish Speech Bank (radio and television broadcasts etc.). In EU network of excellence PASCAL (Pattern Analysis, Statistical Modelling and Computational Learning) I act as a site manager at TKK and I'm the organzinger of MorphoChallenge contest. At the new Adaptive Informatics Research Centre that was recently nominated as one the Centres of Excellence of the Academy of Finland, I'm the head of the Multimodal Interfaces Research Group.
Page maintained by webmaster at cis.hut.fi, last updated Monday, 05-Sep-2011 10:39:36 EEST