Jan van Santen
Phone: 503 346-3765
Fax: 503 346-3754
Director, Center for Spoken Language Understanding
Gaines Hall, Room 102
1979 - Ph.D., Mathematical Psychology, University of Michigan
1974 - M.A., Statistical Psychology, University of Amsterdam
1971 - B.A., Psychology, University of Amsterdam, Amsterdam
Center for Spoken Language Understanding
After obtaining his Ph.D., Dr. van Santen worked at Bell Labs from 1979-1981 and 1984-2000 and at New York University from 1981-1984, on a range of subjects including visual perception and image processing, neuropsychology, and speech technology. His contributions, several of which were patented, consist of a number of quantitative models of various aspects of human behavior, and of associated computational or quantitative methods, including: Measurement of degree of coarticulation in Dysarthria; Small footprint text-to-speech synthesis system; Formant Tracking Based On Phoneme Information; Methods for measuring polysemy based on unannotated text corpora; High-accuracy phonetic feature based speech segmentation; Methods for minimizing spectral discrepancies in acoustic inventories for concatenative synthesis; Statistical modelling of phonetic segment durations via sums-of-products models; Quantitative description of fundamental frequency alignment; Development of "greedy" algorithms for optimal text selection; Automatic phonetic accuracy scoring of orthographic transcriptions of speech; Concept of "command forms" as a user interface for command line based operating systems; Development of a model of clinical judgment in subtyping of developmental language disorders; and Development of a human motion perception model.
Dr. van Santen's current research interests are focused on a broad range of problems in speech and language technology, with special emphasis on basic and applied speech and language technology research for communication disorders.