Abstract

In this paper we present a wearable, headset integrated eyes-and hands-free speech-tospeech (S2S) translation system. The S2S system described here is configured for translingual communication between English and colloquial Iraqi Arabic. It employs an n-gram speech recognition engine, a rudimentary phrase-based translator for translating recognized Iraqi text, and a rudimentary text-tospeech (TTS) synthesis engine for playing back the English translation. This paper describes the system architecture, the functionality of its components, and the configurations of the speech recognition and machine translation engines.

Date

2008

Authors

Kriste Krstovski, Michael Decerbo, Rohit Prasad, David Stallard, Shirin Saleem, Prem Natarajan

Conference

Proceedings of the ACL-08: HLT Workshop on Mobile Language Processing

Pages

10-12