BROWSE
Detail
Conference Paper
An Overview of Korean-English Speech-to-Speech Translation System
Cited - time in 
Share 
Share
- Authors
- Il Bin Lee, Jun Park, Chang Hyun Kim, Young Ik Kim, Sang Hoon Kim
- Issue Date
- 2009-08
- Citation
- Workshop on Technologies and Corpora for Asia-Pacific Speech Translation (TCAST) 2009, pp.1-4
- Language
- English
- Type
- Conference Paper
- Abstract
- During the last 20 years, automatic speech translation has gathered much attention in research community because of its potential to break the language barrier. Speech translation systems for various language pairs are being developed in different countries. In this paper, we present an overview of Korean-English speech-to-speech translation system. The whole system consisted of speech recognition module which was based on finite state transducer and highly improved in terms of its speed, memory consumption, machine translation module enhanced for conversational style, and speech synthesis which can be embedded in handheld devices. The speech translation success rate was measured to evaluate the system's performance, 73.2% from Korean speech to English and 80.3% in the other direction and the results suggested possibility of establishing a commercializable Korean-English speech-to-speech translation system in handheld device.
- KSP Keywords
- Finite state transducer, Handheld Devices, Korean speech, Language barrier, Machine Translation(MT), Speech-to-Speech Translation System, Success rate, memory consumption, speech recognition, speech synthesis, translation module