A New Approach to Speech-Input Statistical Translation

The statistical pattern recognition is a promising framework for text-to-text translation. However, a natural extension to speech-input translation is not straightforward. In this paper, we present a method to deal with the speech input statistical translation problem that could be considered as a step towards a fully integrated recognition-translation procedure. In this version a word graph was used in the input as a representation of the acoustic of a given utterance. As a case study, experimental results with the socalled “Traveller task” are presented by using a text-input statistical translator.

• Type: Conference
• Year: 2000
• Where: ICPR
• Authors: Ismael García-Varea, Alberto Sanchís, Francisco Casacuberta
• Keywords: ICPR 2000, Input Statistical Translation, Statistical Pattern Recognition, Text-input Statistical Translator