Pmtools: A pronunciation modeling toolkit

  title={Pmtools: A pronunciation modeling toolkit},
  author={Richard Sproat},
This paper reports on a pronunciation modeling toolkit — pmtools— that allows one to train a weighted finite-state transducer using a Classification and Regression Tree (CART) training paradigm. Tools are provided to automatically align a pronunciation dictionary consisting of a set of words and their pronunciations, train a set of CART trees on the aligned dictionary, and compile those trees out into a special class of weighted finitestate transducer. Most of the complexity — aligning the data… CONTINUE READING

From This Paper

Figures, tables, and topics from this paper.


Publications referenced by this paper.

Tree-basedmodeling for speechsynthesis

  • M. Riley
  • Bailly andC. Benoit,Eds.Elsevier,
  • 1992
Highly Influential
5 Excerpts

A statisticalmodel for generatingpronunciation networks

  • M. Riley
  • In Proceedingsof theSpeec h andNatural Language…
  • 1991
Highly Influential
4 Excerpts

Multi-level decisiontreesfor staticand dynamicpronunciationmodels

  • E. Fostler-Lussier
  • In Proceedingsof Eurospeec h
  • 1999
1 Excerpt

Stochasticpronunciationmodelling from hand-labelledphoneticcorpora

  • M. Riley, W. Byrne, +7 authors G. andZavaliagkos
  • Speec h
  • 1999
1 Excerpt

Rationaldesignfor aweightedfinite-statetransducer library. LectureNotesin ComputerScience14361998

  • M. Mohri, F. Pereira, M. andRiley
  • 1998
2 Excerpts

Multilingual Text to Speec h Synthesis: TheBell LabsApproach

  • R. Sproat, Ed
  • Kluwer AcademicPublishers, Boston,MA,
  • 1997
1 Excerpt

Similar Papers

Loading similar papers…