Learning Speech Translation from Interpretation
Matthias Paulik · Repository KITopen (Karlsruhe Institute of Technology) · 2010
This thesis introduces methods to directly train speech translation systems on audio recordings of interpreter-mediated communication. By employing unsupervised and lightly supervised training techniques, the proposed methods allow to omit most of the manual transcription effort and all of the manual translation effort that has typically characterized speech translation system development. Thus, the amount of costly and time-consuming human supervision is substantially reduced.