A joint decoding algorithm for multiple-example-based addition of words to a pronunciation lexicon

We propose an algorithm that enables joint Viterbi decoding of multiple independent audio recordings of a word to derive its pronunciation. Experiments show that this method results in better pronunciation estimation and word recognition accuracy than that obtained either with a single example of th...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2009 IEEE International Conference on Acoustics, Speech and Signal Processing s. 4293 - 4296
Hlavní autoři: Bansal, D., Nair, N., Singh, R., Raj, B.
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.04.2009
Témata:
Automatic speech recognition
Bayesian methods
Decoding
Hidden Markov models
Joint decoding
Lattices
Probability distribution
Pronunciation estimation
Speech recognition
US Department of Transportation
Viterbi algorithm
ISBN:9781424423538, 1424423538
ISSN:1520-6149
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:We propose an algorithm that enables joint Viterbi decoding of multiple independent audio recordings of a word to derive its pronunciation. Experiments show that this method results in better pronunciation estimation and word recognition accuracy than that obtained either with a single example of the word or using conventional approaches to pronunciation estimation using multiple examples.
ISBN:9781424423538
1424423538
ISSN:1520-6149
DOI:10.1109/ICASSP.2009.4960578