Speech synthesis apparatus and method
AssigneeKabushiki Kaisha Toshiba
InventorTakehiko KAGOSHIMA
Life of this patent
Estimating the term…
Abstract
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.
Claims9
1-20. (canceled)
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.
A method for storing a plurality of speech unit waveforms corresponding to respective speech units into a waveform memory, comprising:
- inputting a speech unit sequence corresponding to a phoneme sequence of training data;
- counting a frequency of a combination of two speech units continuously appearing in the speech unit sequence;
- selecting the combination of two speech units in higher order of the frequency;
- and
- adjacently storing two speech unit waveforms corresponding to the combination into the waveform memory.
A non-transitory computer readable memory storing program instructions, which when executed by a computer, results in performance of a method for storing a plurality of speech unit waveforms corresponding to respective speech units into a waveform memory, comprising steps of:
- inputting a speech unit sequence corresponding to a phoneme sequence of training data;
- counting a frequency of a combination of two speech units continuously appearing in the speech unit sequence;
- selecting the combination of two speech units in higher order of the frequency;
- and
- adjacently storing two speech unit waveforms corresponding to the combination into the waveform memory.
Dependencies are inferred from the claim text.