Language Technologies

Research in Language Technologies at our lab is mainly motivated by challenges in automatic speech recognition back-end processing such as language modelling, modelling variation in pronunciation, or text alignment. Our text analysis methods comprise customizable phonetic and semantic similarity measures collections with respect to large vocabulary continuous speech recognition (dictation).

Currently, we focus our research on modelling dialect transformations at both, the sentence-level using grammatical transformations and the word-level with pronunciation transformations in close cooperation with our speech synthesis efforts.

Research Area: