2013 · 20 citations · 7 references
ComputingComputer EngineeringComputer ScienceComputational LinguisticsNatural Language ProcessingTranslation TheoryLanguage ModelingLanguage ModelsLexical AnalysisMachine TranslationSpeech RecognitionText GenerationAutomatic Speech TranslationBack-translationN-gram RepresentationN-grams In TranslationSpeech TranslationSpeech-to-text TranslationText-to-speech SynthesisWord N-grams
This work investigates two related issues in the area of WFST-based G2P conversion. The first is the impact that the approach utilized to convert a target word to an equivalent finite-state ma-chine has on downstream decoding efficiency. The second issue considered is the impact that the approach utilized to represent the joint n-gram model via the WFST framework has on the speed and accuracy of the system. In the latter case two novel algorithms are proposed, which extend the work from [1] to enable the use of failure-transitions with joint n-gram models. All solutions presented in this work are available as part of the open-source, BSD-licensed Phonetisaurus G2P toolkit [2]. Index Terms: G2P, WFST, model conversion
7
Conditional and joint models for grapheme-to-phoneme conversion
Stanley F. Chen · 2003 · 145 citations