Title:
SPEECH SYNTHESIZER
Document Type and Number:
Japanese Patent JP3515268
Kind Code:
B2
Abstract:
PROBLEM TO BE SOLVED: To obtain synthesized sounds for which the basic frequencies of phonemes are effectively utilized and which have high distinctness and neutrality by relatively controlling the basic frequencies with are part of the phonemes.
SOLUTION: A basic frequency control section 40 changes the basic frequencies of the phonemes by utilizing the basic frequencies corresponding to part of the phonemes stored in a phoneme data storage section 60. This phoneme data storage section 60 stores the utilization range information predetermining the basic frequencies of which range of the basic frequency information are to be utilized. In such a case, a symbol string input terminal 1 inputs data velating to pronunciation and the phoneme storage section 60 stores the plural phonemes. In the case the basic frequency control section 40 changes the plural basic frequencies possessed by the phonemes to the prescribed basic frequency, the basic frequencies are changed by utilizing the basic frequency corresponding to part of the phonemes and the results thereof are outputted. A speech synthesizing section 5 synthesizes the speeches by coupling the plural phonemes in accordance with the output of the basic frequency control section 40.
Inventors:
Hara, Kiyo
Matsui, Kenji
Kamai, Takahiro
Matsui, Kenji
Kamai, Takahiro
Application Number:
JP5016196A
Publication Date:
April 05, 2004
Filing Date:
March 07, 1996
Export Citation:
Assignee:
MATSUSHITA ELECTRIC IND CO LTD
International Classes:
G10L13/02; G10L13/06; G10L13/07; (IPC1-7): G10L13/08; G10L13/06
Attorney, Agent or Firm:
松田 正道
Previous Patent: DEVICE FOR LEARNING MULTI-LAYER NEURAL NETWORK
Next Patent: MAGNETOSTRICTIVE UNDERWATER VIBRATING DEVICE
Next Patent: MAGNETOSTRICTIVE UNDERWATER VIBRATING DEVICE