A trainable approach for multi-lingual speech-to-speech translation system
Yuqing Gao, Jeffrey S. Sorensen, Hakan Erdoğan, Ruhi Sarikaya, Faju Liu, Michael Picheny, Bowen Zhou, Zhenyu Diao · 2002
This paper presents a statistical speech-to-speech machine translation (MT) system for limited domain applications using a cascaded approach. This architecture allows for the creation of multilingual applications. In this paper, the system architecture and its components, including the speech recognition, parsing, information extraction, translation, natural language generation (NLG) and text-to-speech (TTS) components are described. We have implemented the described system for translating speech between Mandarin and English language pair in an air travel application domain. We are current porting the system to the military domain. Encouraging experimental results have been observed and are presented.