Speech synthesis apparatus and method utilizing acquisition of at least two speech unit waveforms acquired from a continuous memory region by one access
A waveform memory that stores a plurality of speech unit waveforms corresponding to respective speech units, wherein an address order of the speech unit waveforms is determined by a sort order of speech units included in a speech unit sequence corresponding to a phoneme sequence of training data, and the speech units included in the speech unit sequence are selected so as to synthesize a speech of the phone sequence.
1. An apparatus for synthesizing a speech, comprising:
processing circuitry configured to
acquire, from a waveform memory, speech unit waveforms corresponding to a plurality of speech units for synthesizing the speech, and
synthesize the speech by utilizing the speech unit waveforms,
wherein the processing circuitry copies at least two speech unit waveforms from a continuous region of the waveform memory onto a buffer by one access, a data quantity of the at least two speech unit waveforms being less than or equal to a size of the buffer, the at least two speech unit waveforms corresponding to at least two speech units included in the plurality of speech units.
2. The apparatus according to claim 1 , wherein
the processing circuitry selects the plurality of speech units for synthesizing the speech by referring to speech unit information stored in an information memory.
3. The apparatus according to claim 2 , wherein
an access speed of the information memory is quicker than an access speed of the waveform memory.
4. The apparatus according to claim 1 , further comprising:
a text input unit by which a text to be converted to the speech is input.
5. The apparatus according to claim 1 , further comprising:
at least one of a speaker or a head phone by which the speech is output.
6. The apparatus according to claim 1 , further comprising:
a computer operation system configured to execute copying of the at least two speech unit waveforms by the processing circuitry.
7. The apparatus according to claim 1 , further comprising:
a middle-ware (MW) program configured to execute copying of the at least two speech unit waveforms by the processing circuitry.
8. The apparatus according to claim 1 , wherein
the speech unit waveform is one of a waveform itself, a parameter to generate the waveform, encoded data of the waveform, or a pitch waveform.
9. The apparatus according to claim 1 , wherein
the processing circuitry synthesizes the speech by concatenating the speech unit waveforms with modification or without modification.
10. A method for synthesizing a speech, comprising:
acquiring by processing circuitry, from a waveform memory, speech unit waveforms corresponding to a plurality of speech units for synthesizing the speech; and
synthesizing by the processing circuitry, the speech by utilizing the speech unit waveforms,
wherein the acquiring includes
copying at least two speech unit waveforms from a continuous region of the waveform memory onto a buffer by one access, a data quantity of the at least two speech unit waveforms being less than or equal to a size of the buffer, the at least two speech unit waveforms corresponding to at least two speech units included in the plurality of speech units.
11. The method according to claim 10 , further comprising:
selecting by the processing circuitry, the plurality of speech units for synthesizing the speech by referring to speech unit information stored in an information memory.
12. The method according to claim 10 , wherein
the copying is executed by a computer operation system.
13. The method according to claim 10 , wherein
the copying is executed by a middle-ware (MW) program.
14. The method according to claim 10 , further comprising:
outputting the speech by at least one of a speaker or a head phone.
15. The method according to claim 10 , wherein
the speech unit waveform is one of a waveform itself, a parameter to generate the waveform, encoded data of the waveform, or a pitch waveform.
16. The method according to claim 10 , wherein
the synthesizing includes
synthesizing the speech by concatenating the speech unit waveforms with modification or without modification.
17. A non-transitory computer readable memory storing program instructions, which when executed by a computer, results in performance of a method for synthesizing a speech, the method comprising:
acquiring by processing circuitry, from a waveform memory, speech unit waveforms corresponding to a plurality of speech units for synthesizing the speech; and
synthesizing by the processing circuitry, the speech by utilizing the speech unit waveforms,
wherein the acquiring includes
copying at least two speech unit waveforms from a continuous region of the waveform memory onto a buffer by one access, a data quantity of the at least two speech unit waveforms being less than or equal to a size of the buffer, the at least two speech unit waveforms corresponding to at least two speech units included in the plurality of speech units.
18. The non-transitory computer readable medium according to claim 17 , further comprising:
selecting by the processing circuitry, the plurality of speech units for synthesizing the speech by referring to speech unit information stored in an information memory.
19. The non-transitory computer readable medium according to claim 17 , wherein
the speech unit waveform is one of a waveform itself, a parameter to generate the waveform, encoded data of the waveform, or a pitch waveform.
20. The non-transitory computer readable medium according to claim 17 , wherein
the synthesizing includes
synthesizing the speech by concatenating the speech unit waveforms with modification or without modification.