JPH08211896A - System and device for editing speech synthesis - Google Patents

Abstract

PURPOSE: To provide system and device for editing speech synthesis capable of finely controlling the speech synthesis and outputting various and high quality synthesized voices by providing language information obtained by analysis to an acoustic processing part without omitting. CONSTITUTION: This method is a speech synthesis editing method editing the speech synthesis based on sentence information analyzed to accent phrase unit information, and it is provided with a rhythm control process S302 controlling rhythm based on the accent phrase unit information and a generation process S303 generating a synthesized speech parameter based on the rhythm controlled by the rhythm control process.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a speech synthesis editing method and apparatus thereof, and more particularly to a speech synthesis editing method and apparatus for performing speech synthesis editing based on sentence information decomposed into accent phrase unit information.

[0002]

2. Description of the Related Art Generally, in a voice synthesizing device, acoustic information called phonetic text represented by text is input to an acoustic processing unit. In order to generate this phonetic text, sentence analysis processing such as morphological analysis and phrase relation analysis is required.

Using this sentence analysis result, generation and reading of pause information (how much voice is uttered at once (pause phrase) and length of silent time between pause phrases) It performs speech and language processing to generate synthetic speech based on information obtained from language analysis, such as response to changes) and accent generation (combination of accents), and outputs the result as phonetic text. Figure 1 An example of phonetic text
Shown in

In the figure, H85, T300, etc.
The symbols and numerical values enclosed in {} "are the control signal and its value. The alphabets inside the enclosed [[]" are phoneme strings, and "/" and "\" are the rising and falling points of accents. The prosody control unit analyzes (interprets) this phonetic text and extracts information such as accent information, pose information, and pitch frequency necessary for generating prosody information from the phonetic text. Information is "/" and "
From \ ", the pose information is from T300 to 300 (milliseconds),
The pitch frequency is obtained as H85 to 85 (Hz). The prosody control unit generates a pitch pattern from these pieces of information. (See FIG. 6) As described above, acoustic information such as phoneme information, accent information, and control information is written in the phonetic text.

[0005]

However, even if the user corrects the phonetic text to create a synthesized voice of his / her preference, the language information is scarcely written on the phonetic text. Information that can be changed or modified) is limited. Moreover, even if an attempt is made to control prosody using language information, the language information is already missing. If the linguistic information were to be displayed on the phonetic text, the phonetic text would be difficult to read and difficult to understand.

The present invention solves the above-mentioned drawbacks of the related art, and rather than generating phonetic text from sentence analysis results, provides language information obtained by analysis to the acoustic processing section to make speech synthesis more detailed. An object of the present invention is to provide a speech synthesis editing method and a device thereof which can be controlled and can output various and high quality synthesized speech.

[0007]

In order to achieve the above object, a speech synthesis / editing method and apparatus of the present invention have the following configurations. That is, a speech synthesis editing method for performing speech synthesis editing based on sentence information decomposed into accent phrase unit information, the prosody control step of controlling prosody based on the accent phrase unit information, and the prosody control. A generation step of generating a synthetic speech parameter based on the prosody controlled in the step.

Another aspect of the present invention is a speech synthesis editing apparatus for performing speech synthesis editing based on sentence information decomposed into accent phrase unit information, wherein prosody is controlled based on the accent phrase unit information. Prosody control means and generation means for generating synthetic speech parameters based on the prosody controlled by the prosody control means.

[0009]

In the above structure, the prosody is controlled based on the accent phrase unit information, and the synthetic speech parameter is generated based on the prosody controlled in the prosody control step.

According to another aspect of the invention, the prosody control unit controls the prosody based on the accent phrase unit information, and the generation unit generates the synthetic speech parameter based on the controlled prosody.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS First, one of the points of a voice structuring apparatus according to an embodiment of the present invention will be summarized and then a detailed description will be given. The speech structuring apparatus according to the present exemplary embodiment includes an input holding unit that receives a speech language processing result having a frame structure as shown in FIG. 4 as an input, and a prosody that generates prosody control information such as pitch, duration, and power. A control unit, a prosody information holding unit that holds prosody information, a phoneme parameter generation unit that generates synthetic phonetic parameters such as phoneme, parameter connection, or waveform connection; a voice parameter holding unit that holds voice parameters; Basically, a DA converter for converting voice parameters into synthetic voice is provided.

With this basic structure, the output of sentence analysis is expressed by a frame structure described later, and fine control can be performed based on the data of this frame structure, and various synthetic speech outputs are possible. Hereinafter, the audio component device of the present embodiment will be described in detail with reference to the drawings. FIG. 2 is a block diagram showing the configuration of the speech synthesizer of this embodiment.
In the figure, 201 is an input holding unit for inputting and holding frame structure data which is a sentence analysis result, and 202 is
Input frame structure data from the input holding unit 201,
It is a prosody control unit that controls prosody, pitch, duration, power, and the like. Reference numeral 203 denotes a prosody information holding unit that holds prosody control information. A voice generation unit 204 performs processing such as prosody, parameter connection, and waveform connection. Two
Reference numeral 05 is a voice parameter holding unit that holds parameters for generating synthetic voice. Reference numeral 206 denotes a DA converter that converts a voice parameter into synthetic voice.

Next, the operation of this apparatus will be described with reference to the flow chart shown in FIG. First, step S301
Then, it is determined whether or not there is an input of the speech language processing result represented by the frame structure. Here, the frame structure means a structure as shown in FIG. The frame structure is
A collection of pairs of keys and their values.
The key is frame number, notation, accent type, etc., and the respective values are "1", "today" and "1". The value can also point to the next frame by using the key as a pointer (corresponding to the arrow in FIG. 4). The frame in this figure is based on accent phrases. An accent phrase is a unit that has one accent nucleus.

The "speech notation" is a notation of a phoneme string uttered when synthetic speech is generated in katakana.
The "accent type" represents the position of the accent nucleus by the number of mora from the beginning of each accent phrase. Here, the "number of mora" indicates the total number of vowels, consonants, and vowels included in the accent phrase. Further, the “part-of-speech” is a list of the parts-of-speech of the morphemes forming the accent phrase. Furthermore, the “pause length” represents the length between silent phrases generated after the accent phrase.

Returning to FIG. 3, if there is an input in step S301, the process proceeds to step S302. If there is no input, step S301 is repeated. In step S302, the prosody control unit 201 generates prosody control information such as prosody control, pitch generation, and duration. The generated prosody information is held in the prosody information holding unit 203, and step S
Move to 303.

Details of the prosody control method in this step will be described below. FIG. 7 shows a control flow of the prosody control unit 201, and each step will be described below. FIG. 8 shows a detailed configuration of the prosody control unit 201. Here, it is assumed that the leading element of the input accent phrase frame is set in the accent phrase information holding unit 801 in step S301.

In step S701, the accent phrase existence determining unit 802 determines whether or not an accent phrase is set in the accent phrase information holding unit 801. If it does not exist, the process ends. Also, if present,
Then, the process proceeds to step S702. In step S702, the noun determination unit 803 determines whether or not there is a noun in the morphemes forming the accent phrase. If it does not exist, the process proceeds to step S704. If it exists, step S70
Move to 3.

In step S703, prosody information is generated. The pitch frequency setting unit 807 adds a width of +10 to the difference between the maximum value and the minimum value of the pitch frequency given as the default value. That is, this increases the amplitude value of the pitch frequency and produces a pseudo enhancement effect. Further, in order to increase the audibility, the phoneme time length setting unit 808 extends the phoneme time length, and the power setting unit 80
At 9, the magnitude of power is adjusted. Then, the frame next to the accent phrase is set in the accent phrase information holding unit 801, and the process returns to step S701.

Hereinafter, steps S704 to S7
Up to 10, processing is performed to change the pitch frequency, phoneme time length, and power for each part of speech of each constituent morpheme. Step S
In 704, the verb determination unit 803 determines whether or not there is a verb in the morphemes forming the accent phrase.
If it does not exist, the process proceeds to step S706. When it exists, it moves to step S705.

In step S705, prosody information is generated. The pitch frequency setting unit 807 adds a width of +30 to the difference between the maximum value and the minimum value of the pitch frequency given as the default value. That is, this increases the amplitude value of the pitch frequency and produces a pseudo enhancement effect. Further, in order to increase the audibility, the phoneme time length setting unit 808 extends the phoneme time length, and the power setting unit 80
At 9, the magnitude of power is adjusted. Then, the frame next to the accent phrase is set in the accent phrase information holding unit 801, and the process returns to step S701.

In step S706, the number determination unit 805.
Determines whether or not there is a numeral in the morpheme that constitutes the accent phrase. If it does not exist, step S70.
Go to 8. When it exists, it moves to step S707. In step S707, prosody information is generated. The pitch frequency setting unit 807 adds a width of +10 to the difference between the maximum value and the minimum value of the pitch frequency given as the default value. That is, this increases the amplitude value of the pitch frequency and produces a pseudo enhancement effect. Further, in order to increase the audibility, the phoneme duration setting unit 808 extends the phoneme duration and the power setting unit 809 adjusts the power level. Then, the frame next to the accent phrase is set in the accent phrase information holding unit 801, and the process returns to step S701.

In step S708, the particle determination unit 806
Determines whether or not there is a particle in the morphemes that make up the accent phrase. If it does not exist, step S71
Move to 0. When it exists, it moves to step S709. In step S709, prosody information is generated. The pitch frequency setting unit 807 adds a width of -10 to the difference between the maximum value and the minimum value of the pitch frequency given as the default value. That is, this reduces the amplitude value of the pitch frequency. The phoneme duration setting unit 808 shortens the phoneme duration, and the power setting unit 809 adjusts the power level. Then, the frame next to the accent phrase is set in the accent phrase information holding unit 801, and step S7 is performed.
Return to 01.

In step S710, prosody information is generated. In the pitch frequency setting unit 807, the difference between the maximum value and the minimum value of the pitch frequency given as the default value is used as it is. The phoneme duration setting unit 808 sets the phoneme duration as a default value, and the power setting unit 809 sets the default power magnitude. Then, the frame next to the accent phrase is stored in the accent phrase information holding unit 801.
, And the process returns to step S701.

Appropriate prosody control information can be generated by the processing of each step described above. Next, returning to FIG. 3, the processing from step S303 will be described. In step S303, the voice generation unit 204 uses the prosody information stored in the prosody information storage unit 203 to generate phonemes, generate waveforms from parameters, or connect waveforms. The voice signal holding unit 2 stores the generated voice synthesis signal.
The value is held at 05 and the process proceeds to step S304.

In step S304, the voice parameter information is DA-converted and converted into an actual synthesized voice. When the synthetic voice is uttered, all the processes are ended. FIG. 5 shows the configuration of an embodiment of the present invention. 501 is a control memory,
A control program according to the control procedure as shown in the flowchart of FIG. 3 is stored. Reference numeral 502 is a central processing unit for making judgments / calculations and the like according to the control procedure held in the control memory 501. Reference numeral 503 denotes a memory, which is an input holding unit 201, a prosody control unit 202, a prosody information holding unit 20.
3, it has a voice generation unit 204 and a voice signal holding unit 205. Reference numeral 504 is a disk device, and 505 is a bus. (Other Examples) 1. In the above embodiment, the case where the voice parameter generation is performed after the prosody control in FIG. 3 is described, but the present invention is not limited to this, and the prosody control and the voice parameter generation may be performed in parallel. Also, 2. In the above embodiment, the case where the frame structure is used as the input has been described, but the present invention is not limited to this, and any input format having a similar structure may be used. 3. In the above embodiment, the example of prosody generation in the case where the accent component and the natural effect component are separated has been described, but the present invention is not limited to this, and each accent type,
Pitch pattern data may be prepared for each mora number. 4. In the above embodiment, the case where prosody information is generated by the constituent part-of-speech has been described. However, the present invention is not limited to this.
The prosody information may be generated using a factor such as the number of mora.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of a single device. Further, it goes without saying that the present invention can be applied to the case where it is achieved by supplying a program to a system or an apparatus. As described above, according to the present invention, since not the phonetic text but the data of the frame structure is input, it is possible to handle more information than ever, and it is possible to generate a more detailed synthesized voice. The effect is obtained.

[0027]

As described above, according to the present invention, by providing linguistic information obtained by the analysis to the acoustic processing unit, the voice synthesis can be controlled more finely and various and high quality synthesized voice output can be performed. And

[Brief description of drawings]

FIG. 1 is a block diagram showing an example of phonetic text.

FIG. 2 is a block diagram showing a configuration of an exemplary embodiment of the present invention.

FIG. 3 is an operation flowchart showing a processing procedure of the embodiment of the present invention.

FIG. 4 is a diagram showing an example of a frame structure.

FIG. 5 is a basic configuration diagram of a synthetic voice creation support device according to the present invention.

FIG. 6 is a diagram showing a general pitch pattern generation method.

FIG. 7 is a flowchart showing a flow of processing in a prosody control unit.

FIG. 8 is a diagram showing a configuration of a prosody control unit.

[Explanation of symbols]

 201 input holding unit 202 prosody control unit 203 prosody information holding unit 204 voice generation unit 205 voice parameter holding unit 206 DA converter 501 control memory 502 central processing unit 503 memory 504 disk unit 505 bus 801 accent phrase information holding unit 802 accent phrase existence determination Part 803 Noun existence determination part 804 Sensitive verb presence determination part 805 Numeral presence determination part 806 Particle presence determination part 807 Pitch frequency setting part 808 Phonological time length setting part 809 Power setting part

Claims (38)

[Claims] 1. A voice synthesis editing method for performing voice synthesis editing based on sentence information decomposed into accent phrase unit information, comprising a prosody control step of controlling prosody based on the accent phrase unit information, A step of generating a synthetic speech parameter based on the prosody controlled by the prosody control step. 2. The speech synthesis editing method according to claim 1, wherein the accent phrase unit information includes a part of speech of the accent phrase. 3. The speech synthesis editing method according to claim 1, wherein the accent phrase unit information includes the number of mora of the accent phrase. 4. The speech synthesis editing method according to claim 1, wherein the accent phrase unit information includes an accent type of the accent phrase. 5. The speech synthesis editing method according to claim 1, wherein the accent phrase unit information includes a pause length of the accent phrase. 6. The speech synthesis editing method according to claim 2, wherein the prosody control step controls prosody based on the type of part of speech. 7. The speech synthesis editing method according to claim 6, wherein the prosody control step controls the pitch frequency, which is one of the prosody, based on the type of part of speech. 8. The speech synthesis editing method according to claim 3, wherein the prosody control step controls prosody based on the number of mora. 9. The speech synthesis editing method according to claim 8, wherein the prosody control step controls a pitch frequency, which is one of prosody, based on the number of mora. 10. The speech synthesis editing method according to claim 4, wherein the prosody control step controls a pitch frequency, which is one of prosody, based on the accent type. 11. The speech synthesis editing method according to claim 10, wherein the prosody control step controls prosody based on the accent type. 12. The speech synthesis editing method according to claim 5, wherein the prosody control step controls prosody based on the pause length. 13. The speech synthesis editing method according to claim 12, wherein the prosody control step controls a pitch frequency, which is one of prosody, based on the pause length. 14. The speech synthesis editing method according to claim 1, wherein the prosody is a phoneme duration. 15. The speech synthesis editing method according to claim 1, wherein the prosody is sound power. 16. The speech synthesis editing method according to claim 1, further comprising a conversion step of converting to a synthetic speech based on the synthetic speech parameter generated in the generation step. 17. The speech synthesis editing according to claim 7, wherein the prosody control step performs control to increase the width between the maximum value and the minimum value of the pitch frequency if the part of speech is a noun. Method. 18. The method according to claim 14, wherein in the prosody control step, if the part-of-speech is a noun, the phonological time duration is increased. 19. The speech synthesis editing method according to claim 15, wherein in the prosody control step, if the part of speech is a noun, the sound power is increased. 20. A speech synthesis editing apparatus for performing speech synthesis editing based on sentence information decomposed into accent phrase unit information, wherein a prosody control means for controlling prosody based on the accent phrase unit information, A speech synthesis / editing apparatus comprising: a generation unit configured to generate a synthetic speech parameter based on the prosody controlled by the prosody control unit. 21. The speech synthesis editing apparatus according to claim 20, wherein the accent phrase unit information includes a part of speech of the accent phrase. 22. The accent phrase unit information includes the number of mora of the accent phrase.
A voice synthesis editing apparatus described in 0. 23. The speech synthesis editing apparatus according to claim 20, wherein the accent phrase unit information includes an accent type of the accent phrase. 24. The accent phrase unit information includes a pause length of the accent phrase.
A voice synthesis editing apparatus described in 0. 25. The prosody control means controls prosody based on the type of part of speech.
1. A voice synthesis editing apparatus according to 1. 26. The speech synthesis editing apparatus according to claim 25, wherein the prosody control means controls a pitch frequency, which is one of prosody, based on the type of the part of speech. 27. The prosody control means controls prosody based on the number of mora.
The speech synthesis editing apparatus described in. 28. The speech synthesis editing apparatus according to claim 27, wherein the prosody control means controls a pitch frequency, which is one of prosody, based on the number of mora. 29. The speech synthesis editing apparatus according to claim 23, wherein the prosody control means controls prosody based on the accent type. 30. The speech synthesis editing apparatus according to claim 29, wherein the prosody control means controls a pitch frequency, which is one of prosody, based on the accent type. 31. The prosody control means controls prosody based on the pause length.
The speech synthesis editing apparatus described in. 32. The speech synthesis editing apparatus according to claim 31, wherein the prosody control means controls a pitch frequency, which is one of prosody, based on the pause length. 33. The speech synthesis editing apparatus according to claim 20, wherein the prosody is a phoneme duration time length. 34. The speech synthesis editing apparatus according to claim 20, wherein the prosody is sound power. 35. The voice synthesis / editing apparatus according to claim 20, further comprising a converting unit for converting the voice into synthetic voice based on the synthetic voice parameter generated by the generating unit. 36. The voice synthesis edit according to claim 7, wherein the prosody control means performs control to increase the width between the maximum value and the minimum value of the pitch frequency if the part of speech is a noun. apparatus. 37. The speech synthesis editing apparatus according to claim 33, wherein the prosody control means performs control to lengthen the phoneme duration if the part of speech is a noun. 38. The speech synthesis editing apparatus according to claim 34, wherein the prosody control means controls to increase the sound power if the part of speech is a noun.