JPS6024596A - Voice synthesizer - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (1) Technical Field of the Invention The present invention relates to a speech synthesis device which, when synthesizing speech from a character string, changes the speech from standard synthesis to obtain a sound closer to natural speech.

(2) Prior art and problems In the conventional speech synthesis device, as shown in Fig. 1,
The processing unit CPU reads from the file FL that records the character string to be uttered, simultaneously reads the collective specification information from the utterance information file PFL, and sends the combined signal to the voice synthesizer VCM via the communication control unit NCU. '.

For example, when the character string is ``Yamada'', information is added to the character string information that collectively specifies the overall pronunciation time, pitch, etc. so that it can be systematically synthesized into naturally uttered sounds. In the speech synthesis device VCM, the information is given to the phoneme information setting section VS via the communication control device NCU. Since the standard value of the phoneme information is preset in the standard phoneme information table VST, it is used to control the parameter synthesis unit PMC with the value controlled by the above-mentioned batch designation information. The parameter synthesis section PMC reads the parameter file PMF and applies the parameter time series further controlled by the control information G to the voice synthesis section VC to obtain synthesized speech and send it to a speaker or telephone line.

The character string to be uttered at this time is a kana character string, so
For example, when the word ``ame'' is sent, it is impossible to distinguish between rain, candy, and heaven because they are homophones, and an uncertain voice that can be applied to anything is synthesized.

(3) Purpose of the Invention The purpose of the present invention is to improve the above-mentioned drawbacks, and to provide a speech synthesis device which adds simple means to distinguish between homophones and obtain sounds closer to naturally spoken sounds. be.

(4) Structure of the Invention The structure of the present invention to achieve the above-mentioned object is to provide a speech synthesis device that synthesizes speech parameters using standard phoneme information that is set in advance when a character string to be pronounced is sent. , means for adding pronunciation information to each unit that gives one accent type in the character string; means for checking the presence or absence of the added pronunciation information during speech parameter synthesis; It is also necessary to have a means for changing the

(5) Embodiment of the invention FIG. 2 is a block diagram showing the configuration of an embodiment of the invention.
It is shown in correspondence with FIG. The same reference numerals as in FIG. 1 are the same. In Figure 2, pronunciation information is added not to each sentence or to the entire predetermined sentence, but to each unit to which an accent type is assigned, for example, in the character string Yamada, the pronunciation information is added for each pronunciation character such as ya, ma, da, and a new file is created. Store it in FLE. This utterance information is a coded combination of utterance speed, accent, and tone information (intonation information). When the processing device CPU reads the stored information, in this case, the conventional utterance information file PFL is not used, but the information in the file FLE is sent to the speech synthesis device VCM. In the speech synthesis device VCM, in the phoneme information setting section VS, the utterance information processing section VC determines whether or not the character string has the utterance information added thereto. If it is not added, the standard phoneme information table VST is searched for the incoming character string, and the information is sent as is to the variation synthesis section PMC.

In this case, conventional utterance information is collectively designated as a standard value. If it is a character string to which utterance information has been added, the standard phoneme information table VST is searched to obtain that information. 1
Since information about a unit to which an accent type is given, for example, "ya" is sent, the standard information in table VST is changed by three changing devices.

That is, as for the type of voice, for example, there is a device VCE that emits signals separately for male and female voices, a device that changes the time length between syllables and the next syllable using an inter-syllable connection time length setting device TM, and syllable accent information and tone information. HG is used to change the frequency of each syllable depending on the pitch of the voice. Each value changed to a designated value by these devices is input to the parameter synthesis unit PMC. Thereafter, the operation in the parameter synthesis section is not much different from the conventional one shown in FIG. 1, except that the synthesis is slightly more complicated.

Smooth speech synthesis is possible by repeating the operation for each unit to which an accent type is assigned.

(6) Effects of the Invention In this way, according to the present invention, phonation information is added to each unit to which an accent type is assigned, and this information is used to perform rule synthesis, making it easy to create sounds that are closer to naturally spoken sounds. ,Obtainable.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a conventional speech synthesis device, and FIG. 2 is a block diagram showing the configuration of an embodiment of the present invention. V CM, -1- Speech synthesis device FL, FLE - Character string file vs - - Phonological information setting unit VST - Standard phonological information table PMC - Parameter synthesis unit PMF... Parameter file VC - -・Voice information processing unit v c E-Voice type generation device TM'--Syllable duration generation device HG--Syllable frequency native device Patent applicant Fujitsu Limited Agent Patent attorney Eisuke Suzuki

Claims (1)

[Claims] When a character string to be uttered is sent, a speech synthesis device that synthesizes utterance parameters using preset standard phonological information adds utterance information to each unit of the character string to which one accent type is to be assigned. A speech synthesis device comprising: a means for checking the presence or absence of the added utterance information at the time of speech parameter synthesis; and a means for changing the standard phoneme information when the added utterance information is present.