A Modular Approach for Romanian-English Speech Translation

Automatic speech to speech translation is known to be highly beneficial in enabling people to directly communicate with each other when they do not share a common language. This work presents a modular system for Romanian to English and English to Romanian speech translation created by integrating four families of components in a cascaded manner: (1) automatic speech recognition, (2) transcription correction, (3) machine translation and (4) text-to-speech. We further experimented with several models for each component and present several indicators of the system’s performance. Modularity allows the system to be expanded with additional modules for each of the four components. The resulting system is currently deployed on RELATE and is available for public usage through the web interface of the platform.