JP6583756B1 - Speech synthesis apparatus and speech synthesis method - Google Patents

Abstract

A technique capable of editing synthesized speech in real time is provided. A speech synthesizer includes: a display control unit that displays a speech target of synthesized speech to be synthesized in a predetermined display area; and a synthesis unit that performs speech synthesis of the speech target using specified parameters. The parameter is set according to at least one of an instruction input unit that moves a specific position in a display area selected by operating a prepared device, a coordinate of the specified position, a moving speed, and a moving shape. An editing unit for editing, and the parameter is a parameter related to a mode of expression of the speech target, and is a first parameter that adjusts at least the speed of the speech target and an element other than the speed of the speech target speech And a second parameter for adjusting. [Selection] Figure 5

Description

  The present invention relates to a speech synthesis apparatus and a speech synthesis method.

  As a conventional speech synthesizer, a device that synthesizes speech based on text information to be spoken is known (for example, Patent Document 1). Live activities such as virtual singers performing concerts using this technology and live distribution using video distribution sites are also being carried out.

JP2015-102727A

  Such live shows a sense of unity with the audience by changing the way of singing and speaking according to the environment and audience, bringing the value of live. However, in order to change the way of singing and speaking in real time, complicated editing of parameters and the like is required, and it has been difficult to actually use speech synthesis in live performances. Therefore, a voice synthesis technique that can be edited in real time has been desired.

The present invention has been made to solve the above-described problems, and can be realized as the following forms. According to one aspect of the present invention, a speech synthesizer is provided. The speech synthesizer includes: a display control unit configured to display a speech target of a synthesized speech to be synthesized in a display area predetermined by at least one of a character constituting the speech target and a notation for speaking the character; A speech synthesis unit that synthesizes speech using a designated parameter; an instruction input unit that moves a specific position in the display area selected by operating a prepared device; wherein the one or more including at least the moving speed of the position of the coordinates specified moving speed and the moving shape, includes an editing section for editing the parameter, and a reproduction unit for reproducing the synthesized speech, the display control unit scrolls the onset word object scrolling a predetermined speed, the first parameter the parameter for adjusting the speed of at least speech of the calling language object Includes a motor, the first parameter is the determined in accordance with the difference between the scroll speed and the moving speed, the combining unit, sequentially speech using the parameters that reflects the operation by the device retrieved from the editing unit Synthesis is performed, and the synthesized speech synthesized is sequentially reproduced from the reproduction unit. According to the speech synthesizer of this embodiment, the synthesized speech can be edited and reproduced in real time by tracing the speech target, for example. The present invention can be realized in the following forms.

(1) According to an aspect of the present invention, a speech synthesizer is provided. The speech synthesizer includes: a display control unit that displays a speech target of synthesized speech to be synthesized in a predetermined display area; a synthesis unit that performs speech synthesis of the speech target using a specified parameter; An instruction input unit for moving a specific position in the display area selected by operating a prepared device; and the parameter according to at least one of coordinates of the specified position, a moving speed, and a moving shape An editing unit that edits; the parameter includes a first parameter that adjusts at least a speed of the speech to be spoken. According to the speech synthesizer of this form, for example, the synthesized speech can be edited in real time by tracing the speech target.

(2) In the speech synthesizer of the above aspect, the synthesizer may perform speech synthesis using an acoustic model obtained by learning acoustic parameters of the synthesized speech by a statistical method. According to the speech synthesizer of this embodiment, synthesized speech can be generated with a small amount of data.

(3) The speech synthesizer according to the above aspect further includes a reproduction unit that reproduces the synthesized speech, and the synthesis unit sequentially synthesizes speech using a parameter that reflects the operation by the device acquired from the editing unit. And the synthesized speech synthesized may be reproduced from the reproduction unit. According to the speech synthesizer of this form, the synthesized speech can be edited and reproduced in real time.

(4) In the speech synthesizer of the above aspect, the display control unit may display a pointer, and the instruction input unit may realize the movement of the specific position as the movement of the pointer. According to this form of speech synthesizer, parameters can be edited more visually.

(5) In the speech synthesizer of the above aspect, the device may be a pointing device that moves the pointer displayed on the display control unit. According to this form of the speech synthesizer, parameters can be visually edited.

(6) In the speech synthesizer of the above aspect, the device may include the display area. According to the speech synthesizer of this embodiment, for example, the synthesized speech can be edited using a touch panel that can acquire the coordinates and pressure sensitivity that have been touched.

(7) In the speech synthesizer of the above aspect, the editing unit is a parameter related to the expression mode of the speech target according to pressure sensitivity of the device, and is an element other than the speed of the speech target speech You may edit the 2nd parameter which adjusts. According to the speech synthesizer of this aspect, since a plurality of parameters can be edited simultaneously, the synthesized speech can be edited in real time.

  Note that the present invention can be realized in various modes. For example, a speech synthesis system using the speech synthesizer of this embodiment, a method executed in the information processing apparatus to realize the functions of the speech synthesizer and the speech synthesis system, a computer program, and a server for distributing the computer program The present invention can be realized in the form of a device, a non-temporary storage medium storing the computer program, and the like.

It is explanatory drawing which shows the outline | summary of a speech synthesizer. It is an example of the display area displayed by a display control part. It is a figure which shows an example of the various acoustic parameters modeled with an acoustic model. It is a flowchart showing the synthetic | combination audio | voice reproduction | regeneration process in live using a speech synthesizer. It is a flowchart showing a speech synthesis process. It is the figure which showed an example of the locus | trajectory which the instruction | indication input part traced. It is the figure which showed another example of the locus | trajectory which the instruction | indication input part traced. It is explanatory drawing which shows the outline | summary of the speech synthesizer in 2nd Embodiment. It is a flowchart of the speech synthesis process in 2nd Embodiment. It is an example of the display area displayed by the display control part in 2nd Embodiment.

A. First embodiment:
FIG. 1 is an explanatory diagram showing an overview of a speech synthesizer 100 according to an embodiment of the present invention. The speech synthesizer 100 includes a synthesis unit 10, a playback unit 20, a display control unit 30, an instruction input unit 40, an editing unit 50, and a control unit 60.

  The synthesizing unit 10 includes an acoustic model 11 and a synthesis engine 12. The synthesizer 10 synthesizes speech based on the text information to be spoken. In the present embodiment, the synthesis engine 12 performs speech synthesis using the acoustic model 11 in which acoustic parameters are learned by a statistical method and the parameter 13 edited by the editing unit 50 described later. More specifically, a synthesized speech is generated using a hidden Markov model (hereinafter also referred to as HMM (Hidden Markov Model)) or a deep neural network (hereinafter also referred to as DNN (Deep Natural Network)). Details of the acoustic parameters used for learning the acoustic model will be described later.

  The playback unit 20 outputs the synthesized voice generated by the synthesis unit 10 to the speaker 70.

  The display control unit 30 displays the speech target of the synthesized speech in a predetermined display area 31. FIG. 2 is an example of the display area 31 displayed by the display control unit 30. The display area 31 includes display areas 31A, 31B, and 31C that are different from each other. In the present embodiment, the display area 31 has a touch panel 80 which is a device prepared in advance. The touch panel 80 can acquire the coordinates and pressure sensitivity with which the touch pen 85 is in contact. The information that can be acquired by the touch panel 80 is not limited to coordinates and pressure sensitivity, and for example, the tilt of the touch pen 85 may be acquired. In addition, the display control unit 30 is not limited to the touch panel 80 and may be displayed on a normal display or a display included in the speech synthesizer 100. In the present embodiment, for example, speech subject displayed in the display area 31A is "Hello", speech subject displayed in the display area 31B is "thank you", is displayed in the display area 31C speech The subject is "Goodbye".

  The instruction input unit 40 moves a specific position in the display area selected from the display areas 31 </ b> A to 31 </ b> C by operating the touch panel 80 and the touch pen 85 which are devices prepared in advance. The instruction input unit 40 may move a specific position in the selected display area by an operation with a mouse or a finger or an input from a non-contact device such as hand tracking. The specific position may be visually represented by displaying a pointer or the like in the display area or the like. In addition, the movement of the specific position may be represented as a trajectory assuming that the position is being traced with the touch pen 85.

  The editing unit 50 edits the parameters related to the expression mode of the speech target based on the specified position. In this embodiment, the parameters are a first parameter (hereinafter referred to as “speech speed parameter”) that adjusts at least the speed of the speech target, and a volume that is an element other than the speed of the target speech. A second parameter to be adjusted (hereinafter referred to as “volume parameter”) and a third parameter (hereinafter referred to as “pitch parameter”) for adjusting a pitch that is an element other than the speed of the speech to be spoken. . Other parameters other than the speed of speech to be spoken include speech styles such as emotions and habits, vibrato depth and cycle, speaker interpolation ratio, gender parameters (from young feminine voices). Change to the voice of an old man). The parameter preferably includes a speech speed parameter. Further, the utterance target may be changed even during the utterance by moving the position specified across the display areas 31A to 31C.

  The control unit 60 is configured as a computer including a CPU and a memory. The CPU executes the control program stored in the memory, thereby controlling the synthesis unit 10, the playback unit 20, the display control unit 30, the instruction input unit 40, and the editing unit 50 to realize a speech synthesis process described later. .

  FIG. 3 is a diagram illustrating an example of various acoustic parameters modeled by an acoustic model. The fundamental frequency is generally handled as a logarithmic fundamental frequency pt, and as related parameters, voiced / unvoiced discrimination, primary differential (Δpt) and secondary differential (Δ2pt) of the logarithmic fundamental frequency can be considered. These are sometimes called sound source information. The unvoiced part does not have a value of the logarithmic fundamental frequency pt. For this reason, the voiced / unvoiced distinction is made by a method such as putting a predetermined constant in the voiceless part. The spectral parameters include mel cepstrum ct, its first derivative (Δct), and second derivative (Δ2ct). These are sometimes referred to as spectral information. Furthermore, in addition to such sound source information and spectrum information, singing expression information is handled when a singing voice is synthesized.

  The singing expression information includes the vibrato period V1ft and amplitude V1at of the pitch, and the vibrato period V2ft and amplitude V2at of the loudness modeled in units of phonemes and frames. The pitch vibrato period, the pitch vibrato amplitude, the loudness vibrato period, and the loud vibrato amplitude also correspond to the primary differential (Δ) and the secondary differential (Δ2), respectively. However, for the sake of convenience of illustration, in FIG. 3, illustration of the first and second derivatives for the period and the amplitude is omitted. Of the above parameters, the first and second derivatives of each parameter including the mel cepstrum ct are used to take into account time variations. By taking into account the dynamic features, the connection between sounds during sound synthesis becomes smooth. Description of the speech synthesis method using dynamic features is omitted.

  When DNN is used as the acoustic model, a spectrum may be modeled instead of the mel cepstrum ct, or a voice waveform may be modeled as an acoustic parameter instead of the acoustic parameter.

  FIG. 4 is a flowchart showing the synthesized speech reproduction process in live using the speech synthesizer 100 according to the present embodiment. First, the control unit 60 acquires data to be reproduced (step S100). The data to be reproduced is text data to be uttered and musical score information created in advance. The data to be acquired may be directly designated by the user, or may be automatically acquired based on a set list in which a list of songs to be played and utterance contents is listed in order. Note that the processing in step S100 may be omitted, and the speech target may be input on the spot using a keyboard or the like. Next, the control unit 60 controls the instruction input unit 40 to display the data acquired in step S100 (step S110).

  Subsequently, after displaying the data, the control unit 60 performs a speech synthesis process (step S120). The speech synthesis process will be described later. Finally, the control unit 60 determines whether or not the live has ended (step S130). The end of the live can be determined by, for example, whether or not the last music piece and utterance content in the set list have been reproduced. If it is determined that the live has not ended (step S130: NO), the process returns to step S100, and the next data is acquired.

  FIG. 5 is a flowchart showing the speech synthesis process in this embodiment. The voice synthesis process is a process for synthesizing synthesized voice in accordance with the operation of the touch pen 85 by the user. More specifically, it is a process for synthesizing synthesized speech in accordance with the movement of the position specified in the display area 31 by operating the touch pen 85. The instruction input unit 40 first determines whether or not there is an operation with the touch pen 85 (step S200). When there is no operation with the touch pen 85 (step S200: NO), the instruction input unit 40 determines whether or not there is an end instruction (step S240). On the other hand, when there is an operation with the touch pen 85 (step S200: YES), the instruction input unit 40 acquires the movement of a specific position in the display area 31 and the coordinates of the movement locus (step S210).

  The display area 31A will be described as an example with reference to FIGS. The locus is traced from left to right in the direction of the arrow. The synthesizing unit 10 generates a parameter so as to speak in the arrow direction of the locus. Note that when a speech target is specified in the reverse direction, a parameter may be generated so that the speech target is reversely reproduced. As shown in FIG. 6, the locus L1 starts from a position about 1/3 from the left with respect to the substantially rectangular lateral width of the display area 31A in which “ko” is displayed. The editing unit 50 generates a parameter so as to be played back from a portion that has been played back about 1/3 from the start of the utterance with respect to a predetermined length of “ko”. In addition, as shown in FIG. 7, when the instruction input unit 40 traces the locus L3 after tracing the locus L2, the editing unit 50 specifies the speech target specified by the instruction input unit 40 tracing in one direction. A parameter is generated so that “on” and “chiha” are spoken.

  Next, the editing unit 50 edits the parameters of the synthesized speech from the coordinates of the trajectory acquired in step S210 (FIG. 5, step S220). In the present embodiment, the editing unit 50 determines an utterance target for editing the parameter according to the coordinates of the specified position, determines a speech speed parameter according to the moving speed of the specified position, and determines the position of the specified position. A pitch parameter is determined according to the moving shape, and a volume parameter is determined according to the strength of contact (pressure sensitivity) for operating the touch panel 80 of the touch pen 85. More specifically, when the moving speed of the specified position is fast, the speech speed parameter is set so that the speech is uttered in a shorter time than when it is slow. In addition, when the moving shape of the specified position moves upward in the vertical direction of the display area 31, the pitch is raised, and when moved downward, the pitch is lowered. The parameter is set. Further, when the pressure sensitivity of the touch panel 80 is strong, the volume parameter is set so that the speech is uttered at a louder volume than when the touch panel 80 is weak. Note that the assignment of each parameter is not limited to the above-described example. For example, the editing unit 50 may determine the speech speed parameter and the volume parameter according to the movement shape of the specified position. Further, in the present embodiment, the editing unit 50 sounds more sound when the upper part in the vertical direction of the display area 31 is selected than when the lower part is selected, regardless of the moving speed and moving shape of the specified position. Set the height higher.

  Subsequently, the synthesizer 10 synthesizes and reproduces the synthesized speech using the parameters edited in step S220 (step S230). Finally, the synthesis unit 10 determines whether or not there is an end instruction (step S240). For example, when the end instruction is an end instruction that is a press of an end button by the user (step S240: YES), the synthesis unit 10 ends the speech synthesis process. On the other hand, when there is no end instruction (step S240: NO), the synthesis unit 10 returns to the process of step S200.

  According to the speech synthesizer 100 of the present embodiment described above, since the speech speed parameter, the pitch parameter, and the volume parameter can be edited simultaneously by tracing the speech target, the synthesized speech can be edited in real time.

  In the present embodiment, the synthesis unit 10 performs speech synthesis using an acoustic model in which acoustic parameters are learned by a statistical method. Therefore, synthesized speech can be generated without recording speech for each utterance target.

  In the present embodiment, the editing unit 50 determines the volume parameter according to the pressure sensitivity for operating the touch panel 80 of the touch pen 85. Therefore, the loudness can be edited sensuously.

B. Second embodiment:
FIG. 8 is an explanatory diagram showing an overview of the speech synthesizer 100A according to the second embodiment. The speech synthesis apparatus 100A of the second embodiment shown in FIG. 8 is different from the first embodiment in that it includes a storage unit 90 in which accompaniment is stored, and the other configurations are the same. The accompaniment stored in the storage unit 90 is read by the playback unit 20 and output to the speaker 70. The accompaniment memorize | stored in the memory | storage part 90 is the performance data of the accompaniment musical sound produced by the MIDI specification, for example.

  FIG. 9 is a flowchart of speech synthesis processing in the second embodiment. The voice synthesis process of the second embodiment shown in FIG. 8 is different from the first embodiment in that the voice synthesis process is started at the same time as the accompaniment is reproduced and the synthesized voice is a singing voice. Since the configuration of the speech synthesizer of the second embodiment is the same as that of the speech synthesizer of the first embodiment, description of the configuration of the speech synthesizer is omitted.

  The speech synthesizer 100 first reproduces the accompaniment stored in the storage unit 90 (step S300). The timing of starting the speech synthesis processing in the second embodiment may be, for example, when the user presses a playback button, and the speech synthesis is played simultaneously with the accompaniment playback. The accompaniment playback speed is preferably a predetermined playback speed regardless of the editing of the synthesized speech described later. Note that the accompaniment playback timing is associated with the speech target playback timing in advance, so that the playback speed and playback timing are changed according to the change in the speech speed parameter of the synthesized speech, and playback is performed in synchronization with the synthesized speech. Also good. Also, a moving image may be played instead of the accompaniment. Next, the instruction input unit 40 determines whether or not there is an operation with the touch pen 85 (step S310). When there is no operation with the touch pen 85 (step S310: NO), the instruction input unit 40 determines whether or not there is an end instruction (step S350). On the other hand, when there is an operation with the touch pen 85 (step S310: YES), the instruction input unit 40 acquires the coordinates of the specific position in the display region 31, the movement of the specific position, and the coordinates of the movement locus (step S320). .

  FIG. 10 is an example of a display area 31D displayed by the display control unit 30 in the second embodiment. In the present embodiment, the speech target is “Saisai”. The display control unit 30 of the present embodiment may display the pointer P, and the instruction input unit 40 sets the movement of the specific position as the movement of the pointer P. The touch pen 85 is a pointing device that moves the pointer P. The speech target in the display area 31D is scrolled to the left (in the direction of the arrow) at a speed corresponding to the tempo simultaneously with the reproduction of the synthesized speech. The reference position p0 is a reproduction position when reproduction is performed at a predetermined tempo, and is fixed. Note that the width of the content displayed in the display area 31D may be changed by setting the scroll speed to a constant speed regardless of the tempo.

  Next, the editing unit 50 edits the parameters of the synthesized speech from the movement and coordinates of the pointer P acquired in step S320 (FIG. 9, step S330). For example, the editing unit 50 edits the parameter in accordance with the following operation of the pointer P.

<Operation 1> Pointer P selects a specific position <Operation 2> Pointer P traces display area 31D to the left <Action 3> Pointer P traces display area 31D to the right

  In the case of the operation 1 described above, the editing unit 50 determines the playback position according to the position specified by the pointer P with respect to the reference position p0 (hereinafter, specified position p1). More specifically, when the specific position p1 is shifted to the right from the reference position p0, the reproduction position is set so as to seek first according to the magnitude of the shift. The playback may be performed by moving directly from the current playback location to the specific position p1, and the parameters may be set so that the speech target is played back smoothly by interpolating from the current playback location to the specific location p1. Further, the speech target between the current playback location and the specific position p1 may be played back discontinuously.

  In the case of the operation 2 described above, when the pointer P traces the display area 31D in the left direction slower than the scroll speed of the speech target, the editing unit 50 speaks so that the speech is uttered in a time longer than a predetermined tempo. Determine the speed parameter. In addition, the speech speed parameter is set so that reverse playback is performed when the speed is traced faster than the speech target scroll speed. That is, the speech speed parameter is determined by the difference between the scroll speed of the speech target and the speed that the pointer P traces leftward.

  In the case of operation 3 described above, when the pointer P traces the display area 31D to the right, the editing unit 50 sets the speech speed parameter so that the speech is uttered in a time shorter than a predetermined tempo according to the tracing speed. Determine.

  Subsequently, the synthesizer 10 synthesizes synthesized speech using the parameters edited in step S330 and reproduces them (step S340). The synthesizer 10 determines whether or not there is an end instruction (step S350). The end instruction is, for example, pressing of an end button by the user. When there is no end instruction (step S350: NO), the synthesizer 10 determines whether or not the accompaniment reproduction has been completed (step S360). If reproduction has not been completed (step S360: NO), the processing is repeated from step S310. When the reproduction is completed (step S360: YES), the speech synthesis process ends. Even when it is determined in step S360 that the accompaniment reproduction has been completed, if the input to the instruction input unit 40 continues, the synthesis unit 10 continues to output the synthesized speech until there is no input to the instruction input unit 40. You may continue playing. For example, reproduction may be performed such that the last lyrics of the synthesized speech are extended with a long tone.

  In the present embodiment, while the input to the instruction input unit 40 is not performed, the reproduction unit 20 reproduces the synthesized voice with mute until input to the instruction input unit 40 again. Speech synthesis may be performed so that playback is performed using the speech speed parameter. The synthesizing unit 10 may perform speech synthesis so as to stop the playback of the synthesized speech. For example, when there is no input to the instruction input unit 40 for a predetermined time, the synthesis unit 10 fades out so as to stop the reproduction of the synthesized speech before the planned synthesis speech reproduction is completed. The speech synthesis may be performed as described above, and then the playback of the synthesized speech may be stopped. Further, the reproduction unit 20 may reproduce the sound with mute before the planned reproduction of the synthesized voice is completed. Note that the accompaniment may continue to be played, or may be played back with a pause or mute simultaneously with the synthesized voice.

  According to the speech synthesizer 100 of the present embodiment described above, the speech speed parameter can be edited by tracing the speech object being reproduced, so that the synthesized speech can be edited and reproduced in real time.

C. Other embodiments:
In the above embodiment, the synthesis unit 10 performs speech synthesis using an acoustic model in which acoustic parameters are learned by a statistical method. Instead, the synthesis unit 10 may perform speech synthesis using a waveform connection method.

  In the above embodiment, the synthesizing unit 10 may store the generated parameter value, and at the next playback of the synthesized speech, may perform speech synthesis using the stored parameter value for playback. Here, “reproduction of the synthesized speech” means that the synthesized speech is reproduced after the speech synthesis process is completed.

  In the above embodiment, the editing unit 50 determines parameter values included in the parameters according to the shape, speed, and pressure sensitivity of the trace traced by the instruction input unit 40. Instead, the combining unit 10 may determine parameter values included in each parameter using values of the tilt of the touch pen 85 and values of an external controller that is a device prepared in advance.

  Moreover, in the said embodiment, the display control part 30 may display several utterance area | regions from which a pitch differs corresponding to each part of utterance content and a score. According to this aspect, by selecting each speech area by the instruction input unit 40, it is possible to synthesize various voices while shifting. Further, the input of the instruction input unit 40 to one speech area may be reflected to the other instruction input units 40.

  In the second embodiment, the editing unit 50 determines the playback position according to the specific position p1. Instead, the editing unit 50 may determine the speech speed parameter according to the specific position p1. For example, when the specific position p1 is shifted to the right from the reference position p0, the speech speed parameter may be set so that the speech speed increases according to the magnitude of the shift.

  In the second embodiment, the editing unit 50 determines the speech speed parameter according to the tracing speed of the pointer P, but it may also change the scroll speed of the speech target.

  The present invention is not limited to the above-described embodiment, and can be realized with various configurations without departing from the spirit of the present invention. For example, the technical features in the embodiments corresponding to the technical features in each embodiment described in the summary section of the invention are for solving the above-described problems or achieving some or all of the above-described effects. In addition, replacement and combination can be performed as appropriate. Further, if the technical feature is not described as essential in the present specification, it can be deleted as appropriate.

    DESCRIPTION OF SYMBOLS 10 ... Synthesis | combination part, 11 ... Acoustic model, 12 ... Synthesis engine, 13 ... Parameter, 20 ... Playback part, 30 ... Display control part, 31, 31A, 31B, 31C, 31D ... Display area, 40 ... Instruction input part, 50 ... Editing unit, 60 ... Control unit, 70 ... Speaker, 80 ... Touch panel, 85 ... Touch pen, 90 ... Storage unit, 100, 100A ... Speech synthesizer, L1, L2, L3 ... Track, P ... Pointer, p0 ... Reference position , P1 ... specific position

Claims (10)

A speech synthesizer,
A display control unit that displays a speech target of the synthesized speech to be synthesized in a predetermined display area by at least one of a character constituting the speech target and a notation for speaking the character;
A synthesizer that synthesizes the speech object using a designated parameter;
An instruction input unit for moving a specific position in the display area selected by operating a device prepared in advance;
An editing unit that edits the parameter according to one or more including at least the moving speed among the coordinates, moving speed, and moving shape of the specified position;
A playback unit for playing back the synthesized speech,
The display control unit scrolls the speech object at a predetermined scroll speed,
The parameter includes a first parameter that adjusts at least a speed of the speech of the speech target;
The first parameter is determined according to a difference between the scroll speed and the moving speed,
The speech synthesis apparatus, wherein the synthesis unit sequentially performs speech synthesis using parameters obtained from the editing unit and reflecting the operation by the device, and sequentially reproduces the synthesized speech from the playback unit. The speech synthesizer according to claim 1,
The synthesis unit is a speech synthesizer that performs speech synthesis using an acoustic model obtained by learning acoustic parameters of the synthesized speech by a statistical method. The speech synthesizer according to claim 1 or 2,
The display control unit can display a pointer,
The instruction input unit is a speech synthesizer that realizes movement of the specific position as movement of the position of the pointer. The speech synthesizer according to claim 3,
The speech synthesizer, wherein the device is a pointing device that moves the pointer displayed on the display control unit. The speech synthesizer according to any one of claims 1 to 3,
The speech synthesizer, wherein the device is a touch panel having the display area. The speech synthesizer according to any one of claims 1 to 5,
The editing unit edits a second parameter that adjusts an element other than the speed of the speech target speech according to a pressure sensitivity of the device and adjusts elements other than the speed of the speech target speech. Synthesizer. The speech synthesizer according to any one of claims 1 to 6,
The speech synthesis apparatus, wherein the synthesis unit performs speech synthesis in parallel with the reproduction unit reproducing the accompaniment, and sequentially reproduces the synthesized speech synthesized from the reproduction unit. The speech synthesizer according to any one of claims 1 to 7,
The speech synthesizer, wherein the display control unit displays the utterance target in a display area predetermined by characters constituting the utterance target and a notation for speaking the character. The speech synthesizer according to any one of claims 1 to 8,
The display control unit displays, in the display area, at least one of each character constituting the speech object and each notation for speaking the character according to a length corresponding to the time to speak. A speech synthesizer. A speech synthesis method,
A display step of displaying a speech target of the synthesized speech to be synthesized in a predetermined display area by at least one of a character constituting the speech target and a notation for speaking the character;
A speech synthesis step of synthesizing the speech object using designated parameters;
A moving step of moving a specific position in the display area selected by operating a device prepared in advance;
A scrolling step of scrolling the speech object at a predetermined scrolling speed;
The above one including at least the moving speed of the movement trajectory and the coordinates and the moving speed of the position to be the specific, an editing step for editing the parameters,
In the speech synthesis step, a sequential speech synthesis is performed using parameters reflecting the operation by the device acquired in the editing step, and a playback step of sequentially reproducing the synthesized speech synthesized, and
The parameters observed contains a parameter for adjusting a speed of at least speech of the calling language object, the first parameter is determined according to the difference between the scroll speed and the moving speed, the speech synthesis method.

Images