Using Word Graphs as Intermediate Representation of Uttered Sentences
We present an algorithm for building graphs of words as an intermediate representation of uttered sentences. No language model is used. The input data for the algorithm are the pronunciation lexicon organized as a tree and the sequence of acoustic frames. The transition between consecutive units are considered as additional units.
Nodes represent discrete instants of time, arcs are labelled with words, and a confidence measure is assigned to each detected word, which is computed by using the phonetic probabilities of the subsequence of acoustic frames used for completing the word.
We evaluated the obtained word graphs by searching the path that best matches with the correct sentence and then measuring the word accuracy, i.e. the oracle word accuracy.
Keywordsword graphs word lattices lexical tree confidence measures
- 2.Ney, H., Ortmanns, S., Lindam, I.: Extensions to the word graph method for large vocabulary continuous speech recognition. In: Proceedings of IEEE ICASSP 1997, Munich, Germany, vol. 3, pp. 1791–1794 (1997)Google Scholar
- 4.Ferreiros, J., San-Segundo, R., Fernández, F., D’Haro, L.-F., Sama, V., Barra, R., Mellén, P.: New word-level and sentence-level confidence scoring using graph theory calculus and its evaluation on speech understanding. In: Proceedings of INTERSPEECH 2005, Lisbon, Portugal, pp. 3377–3380 (2005)Google Scholar
- 9.Calvo, M., Gómez, J.A., Sanchis, E., Hurtado, L.F.: An algorithm for automatic speech understanding over word graphs. Procesamiento del Lenguaje Natural (48) (accepted, pending of publication, 2012)Google Scholar
- 10.Moreno, A., Poch, D., Bonafonte, A., Lleida, E., Llisterri, J., Mariño, J.B., Nadeu, C.: Albayzin Speech Database: Design of the Phonetic Corpus. In: Proceedings of Eurospeech, Berlin, Germany, vol. 1, pp. 653–656 (September 1993)Google Scholar
- 11.Benedí, J.M., Lleida, E., Varona, A., Castro, M., Galiano, I., Justo, R., López, I., Miguel, A.: Design and acquisition of a telephone spontaneous speech dialogue corpus in Spanish: DIHANA. In: Proc. of LREC 2006, Genova, Italy (2006)Google Scholar