JPH0546195A - Speech synthesizing device - Google Patents

Abstract

PURPOSE:To synthesize a speech by accessing speech data according the attribute data of the speech data. CONSTITUTION:A memory part 11 is stored with the speech data of plural phrases and the attribute data of the respective speech data. At a retrieval request, the attribute data are read out of the memory part 11 by a retrieval part 12, which compares input attribute data for retrieval with the attribute data read out of the memory part 11 to retrieve the attribute data, and outputs the retrieval result to an addressing part 13. The addressing part 13 generates address data for a speech data storage part corresponding to the retrieval output of the retrieval part 12 and sends this address data to the memory part 11; and the speech data are read out of the memory part 11 and sent to a speech synthesis part 14, which synthesizes the speech.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a speech synthesizer,
In particular, the present invention relates to a voice synthesizing device capable of selecting a voice to be synthesized based on the attribute of the voice phrase when synthesizing the voice.

[0002]

2. Description of the Related Art FIG. 5 is a block diagram of a conventional speech synthesizer capable of synthesizing a plurality of phrases. This voice synthesizer prepares as many data storage devices 31 as the number of phrases, and when synthesizing a voice, one is selected from a plurality of data storage devices 31 and stored in the selected data storage device 31. The voice synthesis is performed by sending the voice data to the voice synthesizer 32.

FIG. 6 is a block diagram of a conventional speech synthesizer different from the above. In this voice synthesizer, an index area and a voice data area are set in a data storage device, for example, an IC memory 33, voice data for a plurality of phrases is stored in the voice data area, and the voice data of these phrases is stored in the index area. The address (AD) is stored, and the voice data is selected by referring to the index region and addressing the voice data region during voice synthesis.
Voice synthesis is performed by sending the selected voice data to the voice synthesis unit 32.

In both of the conventional voice synthesizers, voice data corresponding to a plurality of voice phrases are discretely accessed in any case. Therefore, for example, it is possible to access random audio data in which the order is specified, such as the first from the beginning, the second, etc. However, it is possible to access the audio data by data other than the storage location (address) of the audio data. Can not. That is, there is a problem that the voice data cannot be accessed based on the attribute data of the voice data, such as time data storing the voice data and the data length of the voice data.

[0005]

As described above, the conventional voice synthesizer has a problem that the voice data cannot be accessed based on the attribute data of the voice data.

The present invention has been made in consideration of the circumstances as described above, and an object thereof is to access a voice data based on attribute data of the voice data and perform a voice synthesis. Is to provide.

[0007]

According to the present invention, a voice data storage unit for storing voice data of a plurality of phrases,
An attribute data storage unit for storing attribute data corresponding to each voice data of the plurality of phrases, and a search for voice data stored in the voice data storage unit based on the attribute data stored in the attribute data storage unit Search section,
A voice data selecting unit for selecting voice data in the voice data storing unit based on a search result of the search unit; and a voice synthesizing unit for synthesizing voice from the voice data selected by the voice data selecting unit. Is characterized by.

[0008]

[Operation] The voice data stored in the voice data storage unit is searched based on the attribute data corresponding to the voice data,
The voice data stored in the voice data storage unit is selected based on the search result. As a result, the voice data can be accessed by data other than the storage location of the voice data itself.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the overall structure of a first embodiment of a speech synthesizer according to the present invention. In the figure, 11 is a memory unit for storing voice data for a plurality of phrases and attribute data of each voice data. As the memory unit 11, for example, an IC memory, a flexible disk storage device, a hard disk storage device, a magnetic tape storage device or the like is used. The attribute data referred to here is
This is data for identifying voice data, and includes, for example, the date and time when the voice data was created, the data length of the voice data, the speaker name, and the gender of the speaker. The respective voice data and the attribute data are paired and stored in the memory unit 11 as shown in the figure.

Reference numeral 12 is a search unit. The search unit 12 reads the attribute data stored in the memory unit 11 when a search request is made, and inputs the search attribute data and the memory unit 11.
The attribute data read from 11 is compared to search the attribute data, and the search result corresponding to the attribute data having the matched content is output to the addressing unit 13. The addressing unit 13 generates address data of the voice data storage unit corresponding to the search output of the search unit 12. By sending this address data to the memory unit 11, the voice data is read from the memory unit 11 and sent to the voice synthesis unit 14 to synthesize the voice. It should be noted that PARCOR is used as the voice synthesis unit 14.
(Partial Auto-Correlation) method, ADM (Adaptive D
ifferrential Modulation) method, ADPCM (Adaptive D
ifferrential Pulse Code Modulation) method, MPC (Mu
It is possible to use a waveform synthesizing circuit such as a waveform coding method such as the lti Pulse Cording method and an analysis coding method.

As described above, in the voice synthesizer of the above-described embodiment, the voice data is generated based on the attribute data of the voice data such as the date and time of the voice data creation, the data length of the voice data, the speaker name, and the gender of the speaker. You can access and perform voice synthesis.

FIG. 2 is a block diagram showing the overall construction of a second embodiment of the speech synthesizer of the present invention. The device of this embodiment differs from the device of the first embodiment in that
Instead of the individual memory units 11, the memory unit 11 for voice data is used.
A and a memory section 11B for attribute data are provided so that voice data and attribute data are stored in independent memory sections.

FIG. 3 is a block diagram showing the overall structure of a third embodiment of the speech synthesizer of the present invention. In this embodiment, as the attribute data stored in the memory unit 11, as described above, in addition to data such as the creation date and time of the voice data, the data length of the voice data, the speaker name, the gender of the speaker, and the like, It also contains data for selecting the composition method.

In the apparatus of this embodiment, one of the search data input to the search unit 12 also includes data for selecting a voice synthesis method. Then, this data is supplied to the switch unit 15 to which the audio data read from the memory unit 11 is input. The switch unit 15 uses the voice data read from the memory unit 11 based on the data for selecting the voice synthesis system output from the search unit 12, and a plurality of voice synthesis units having different voice synthesis systems (3 in this embodiment). Speech synthesizer
14A, 14B, 14C). Of these three voice synthesizers 14A, 14B, and 14C, the one to which voice data is supplied synthesizes the voice according to the synthesis method. The three speech synthesis units 14A, 14B, and 14C have the above-mentioned PARCOR scheme and ADM synthesis scheme.
It is possible to select from the methods, ADPCM method, MPC method, etc.

As in the case of the first embodiment apparatus shown in FIG. 1, this embodiment apparatus uses the voice data such as the creation date and time of the voice data, the data length of the voice data, the speaker name, and the gender of the speaker. Access voice data based on data attribute data,
It has an effect that voice synthesis can be performed. In addition to the above effects, the device of this embodiment also has an effect that the memory usage efficiency can be increased because a common memory unit is used for a plurality of voice synthesis units.
Furthermore, since a voice synthesis unit for multiple systems is provided, a voice synthesis system with a good data compression rate (PA
There is also the effect that it is possible to select and use the RCOR method) or the voice synthesis method (ADPCM method) with good sound quality.

FIG. 4 is a block diagram showing the overall construction of the fourth embodiment of the speech synthesizer of the present invention. The apparatus of this embodiment is different from the apparatus of FIG. 3 in that a plurality of voice encoding units (two voice encoding units 16A and 16B in this embodiment) having different methods, an attribute data creation unit 17, and an input. A switch unit 18 for selectively supplying voice to the two voice encoding units 16A and 16B is added.

Here, the attribute data creating section 17 creates attribute data for each input voice, and the attribute data created here is sent to and stored in the memory section 11. Also, the two speech coding units 16A and 16B are
Audio data is created by encoding the input audio based on the audio encoding method selected from the RCOR method, ADM method, ADPCM method, MPC method, and the like. Since the device of this embodiment includes the voice encoding unit, there is an effect that the voice phrase to be synthesized can be rewritten.

[0019]

As described above, according to the present invention,
It is possible to provide a voice synthesizing device that can perform voice synthesis by accessing voice data based on attribute data of voice data. Further, since the voice data can be handled as a processing target of the data file or the database, there is an effect that the voice data can be smoothly introduced into the conventional data processing system.

[Brief description of drawings]

FIG. 1 is a block diagram of an apparatus according to a first embodiment of the present invention.

FIG. 2 is a block diagram of a second embodiment device of the present invention.

FIG. 3 is a block diagram of an apparatus according to a third embodiment of the present invention.

FIG. 4 is a block diagram of an apparatus according to a fourth embodiment of the present invention.

FIG. 5 is a block diagram of a conventional device.

FIG. 6 is a block diagram of a conventional device.

[Explanation of symbols]

11, 11A, 11B ... Memory section, 12 ... Search section, 13 ... Addressing section, 14, 14A, 14B, 14C ... Speech synthesis section, 15, 18
... switch section, 16A, 16B ... voice encoding section, 17 ... attribute data creating section.

Claims (4)

[Claims] 1. A voice data storage unit for storing voice data of a plurality of phrases, an attribute data storage unit for storing attribute data corresponding to each voice data of the plurality of phrases, and an attribute data storage unit stored in the attribute data storage unit. A search unit for searching voice data stored in the voice data storage unit based on attribute data; a voice data selection unit for selecting voice data in the voice data storage unit based on a search result of the search unit; A voice synthesizing device comprising: a voice synthesizing unit for synthesizing a voice from the voice data selected by the voice data selecting unit. 2. The voice synthesizer according to claim 1, further comprising a voice data forming unit which forms voice data from input voice. 3. The voice synthesizer according to claim 1, further comprising an attribute data forming unit that forms attribute data for the voice data stored in the voice data storage unit. 4. The voice data storage section and the attribute data storage section are configured by one memory device.
The described speech synthesizer.