ITWissen.info - Tech know how online

speech-to-speech translation (S2ST)

Speech-to-Speech Translation (S2ST or SST) is speaker-independent bidirectional speech-to-speech translation in real time, during a dialogue.

S2ST translation supports real-time interpersonal communication between people who do not speak the same language. In this translation technique, a speech signal is translated into a source language, which is then translated into the target person's language.

A speech-to-speech system consists of speech recognition, the translation engine, Machine Translation( MT), and subsequent speech synthesis. The machine part receives the speech texts that a speech recognizer recognizes. Since the spoken word differs from the written word, including grammatically, the rule-based Machine Translation cannot translate the texts in the same way as written texts. Therefore, S2ST systems use, among other things, spoken language examples, Example-Based Machine Translation (EBMT), which are stored in a database and used during translation. The translation must be context-based, which is problematic in the case of ambiguity.

Informations:
Englisch: speech-to-speech translation - S2ST
Updated at: 17.09.2012
#Words: 147
Links: simple serial transport (bus) (SST), bidirectional (BiDi), speech, translation, communication (COM)
Translations: DE
Sharing:    

All rights reserved DATACOM Buchverlag GmbH © 2024