CMU Sphinx2 Default Phone Set

for American English. Last Modified: March 12, 2001 by lenzo@cs.cmu.edu


Speech at CMU | Sphinx

The default phone set for CMU Sphinx2 is a DARPA-bet derivative, with noise-phones and without lexical stress. This is similar to, but not precisely the same as, the CMU Pronouncing Dictionary, CMUDICT.

In 2000, the Sphinx 2 default phone set was modified to remove "deletable" stops, and to split the TS "phone" into T and S. The phone set that is used for the dictionary must match the set of phones in the acoustic models.

  1. SIL

  2. AA
  3. AE
  4. AH
  5. AO
  6. AW
  7. AX
  8. AXR
  9. AY
  10. B
  11. CH
  12. D
  13. DH
  14. DX
  15. EH
  16. ER
  17. EY
  18. F
  19. G
  20. HH
  21. IH
  22. IX
  23. IY
  24. JH
  25. K
  26. L
  27. M
  28. N
  29. NG
  30. OW
  31. OY
  32. P
  33. R
  34. S
  35. SH
  36. T
  37. TH
  38. UH
  39. UW
  40. V
  41. W
  42. Y
  43. Z
  44. ZH

  45. +BACKGROUND+
  46. +BREATH+
  47. +COUGH+
  48. +LAUGH+
  49. +NOISE+
  50. +SMACK+
  51. +TONE+
  52. +UH+
  53. +UHUM+
  54. +UM+