JP3404776B2 - Signal playback device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [0001] BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention reproduces sound including sound.
Device that reproduces sound and image information such as map information
To a signal reproducing device. For the sound reproduction of the present invention,
Suitably, a rule combining scheme is used. Signal reproducing device of the present invention
Is used, for example, in guidance information devices for vehicles such as automobiles.
It is preferably applied. [0002] 2. Description of the Related Art Conventionally, devices for reproducing voice and other sounds have been used.
Or the reproduction of map information for such sound reproduction.
Vehicle guidance information device used for vehicle guidance
Has been proposed. In such a vehicle guidance information device,
However, due to various restrictions on mounting a vehicle,
There is a demand for reduced coding and efficient encoding and decoding.
And a high efficiency code is used. [0003] SUMMARY OF THE INVENTION However, the conventional
In vehicle guidance information devices, the degree of reduction in storage capacity
Is only a few kilobits per second (Kbps) at most
Even if high-efficiency coding is applied, a large amount of audio signals can be encoded.
Requires still a significant amount of storage
There is a problem that there is. An object of the present invention is to reproduce sound or reproduce sound.
Possible signal reproduction to reproduce image information such as maps in addition to raw
In storage, the storage capacity of the storage device that stores the signal for reproduction
Is to reduce. The purpose of the present invention is to
An object of the present invention is to provide a signal reproducing apparatus capable of shortening the time. [0005] [MEANS FOR SOLVING THE PROBLEMS] The present invention for solving the above problems.
The signal reproducing apparatus according to the first aspect of the present invention has an analysis processing for reproduction in advance.
Phonetic symbol information and sound for intelligent speech synthesis
Acoustic information including the synthesis unit and corresponding to the acoustic information
Recording means for recording text information;In the above recording means
Limited voice utterances among the recorded phonetic symbol information
Selection information recording means for recording selection information of phonetic information,
Based on the above selection informationThe sound information from the recording means
First reading means for readingReadacoustic
Sound synthesis means for generating a sound reproduction signal from information;
CreatedMeans for reproducing a sound reproduction signal;
Second reading means for reading the text information, and the text
Display means for displaying information.Be preparedCharacterized by
You. [0006] [0007] [0008] [0009] [0010] [0011] For the voiced part of the voice unit data, the actual sound
Extraction of voiced parts using complex cepstrum analysis
Was issued. Impulse and unit response wave corresponding to one pitch
Form as one set, this set as one voice unit data
It consists of only the required pitches stored, and also has phoneme units.
For the unvoiced part of the data, the waveform of the unvoiced part of the real voice
It is cut out and stored as it is. [0013] [Action] The acoustic symbol information obtained from the acoustic text is pre-
Sentences are analyzed and calculated by speech synthesis rule processing.
The voice pronunciation calculated in this way is stored in the storage device of the signal reproduction device.
The symbol information is stored. As a result, the required storage capacity
Can be reduced. [0014] In addition, from the acoustic text to the acoustic symbol, preferably
Is a sentence analyzer and dictionary that converts to phonetic symbol information
Without having to use the phonetic phonetic symbol information stored in this way.
Playback reduces the time required for sentence analysis and plays back
Time can be reduced. However, it also stores speech synthesis segments
You can use this segment to play back audio
You. [0015] In addition to the above audio reproduction, reproduction of images such as maps.
Can also be performed. Also, if the user responds to the audio output
The correspondence between audio output and input when inputting
Wear. [0016] Embodiments of the present invention will be described below in detail with reference to the drawings.
Will be described. FIG. 1 shows a signal reproducing apparatus according to a first embodiment of the present invention.
Sound reproduction that reproduces synthesized sounds from acoustic information such as various voices
It is a block diagram of an apparatus. In this embodiment, the sound
In particular, a case where sound is reproduced will be exemplified. That is,
The first embodiment relates to a device for reproducing sound. Shown in FIG.
The voice reproducing device 101 stores the phonetic symbol information 103.
Recording device 102, voice pronunciation from the recording device 102
A readout circuit 104 for reading out the symbol information 103;
And a voice synthesizer 105. Voice synthesizer 105
An audio output device such as a speaker is connected to the subsequent stage
However, it is not shown. Also, in the device described below
Also, illustration of an audio output device such as a speaker is omitted. In a first embodiment of the present invention, a text
From the sentence analysis unit until the phonetic symbols are obtained
The phonetic transcription that is not built in the device 101 but has been analyzed in advance
Is recorded in the recording device 102 as phonetic symbol information 103.
Record it. In other words, phonetic symbols are pre-
The speech information processed and analyzed in advance by the sentence analysis unit
The information is stored in the storage device 102. As a result, during playback
Since the analysis time is not required, the reproduction time can be reduced.
Details of the analysis will be described later. As the recording device 102, Compact is used.
  Disc (CD) device, Mini-Disc (MD)
Large-capacity, small-sized read-only record as exemplified in devices
A read-only optical disk device as a storage device is suitable.
Alternatively, as the above-mentioned storage device, a recording device such as an MD device
Separate from magneto-optical disk drive represented by raw disk drive
Disk device for recording and playback using various principles, for example,
Disk device such as phase change type optical disk device, ROM
Also uses semiconductor memory such as RAM flash memory
sell. Magneto-optical disk drives use a light spot to
Increase the temperature of the minimal area to make it easy to magnetize,
Multiply by “1” or “0” to store the digital
It is a recording medium. The phase-change type disk drive uses a light spot
A method of recording information by causing a phase change in the metal film by using
It is a recording medium. Flash memory on FET gate
Turns FET on by accumulating or removing electric charge in insulating film
.A recording medium for storing data off, and an area for recording information
Is small and economical, can operate at high speed,
It has features such as the fact that no power supply is
You. However, compared to optical disk devices, the storage capacity is
It has the disadvantage of being less expensive. [0019] stored in the phonetic symbol information 103
Details of offline creation of phonetic symbol information 103
This will be described with reference to FIG. FIG. 2 shows an arithmetic processing unit as a whole.
1 shows a schematic configuration of a speech synthesizer 201 having an arrangement configuration;
The speech synthesizer 201 includes a speech unit storage unit 202, a sentence
Chapter analysis unit 203, speech synthesis rule unit 204, and speech synthesis
Unit 205. Audio reproduction device 10 shown in FIG.
1 is installed in a vehicle and operates online in real time
In contrast, the speech synthesizer 201 shown in FIG.
Phonetic symbols reproduced and processed in advance offline
The information 103 is stored in the recording device 102. The sentence analysis unit 203 receives a message from a predetermined input device.
Entered text input (sentences represented by a series of characters
Etc.)] based on a predetermined "dictionary"
After analyzing the text and converting it to a kana character string, words and phrases
Decompose each time. In Japanese, words are like English
For example, "U.S. industry
Analyzing words like "world", "US / industry / world"
It can be divided into two types, such as "US / Domestic / Industry". Sentence
The chapter analysis unit 203 refers to the “dictionary” and
Utilize continuity and statistical properties of words
For example, if you enter the word "US industry"
Break down each word and phrase, and use this to determine the boundaries of words and phrases.
To detect. The sentence analysis unit 203 thus performs, for each word,
After detecting basic accents, sound the detected
Output to the composition rule unit 204. [0021] The speech synthesis rule unit 204 provides
According to the predetermined "phonological rules" set based on the sentence,
Processes the above detection result and text input of the analysis unit 203
I do. Japanese natural speech is based on linguistic characteristics
To distinguish them, divide them into about 100 utterance units
Can be. For example, the word "Sakura"
When divided into three places, three CVs of "sa" + "ku" + "ra"
Can be divided into units. Furthermore, in Japanese, the words
If they are consecutive, the initial syllable of the succeeding word becomes muddy.
(Ie, it becomes turbid)
It is said that the utterance changes from the case of a single word due to sound
There is a feature. The speech synthesis rule unit 204 uses these Japanese
Processing so that "phonological rules" are set according to the characteristics of
In accordance with the rules, the text
The input is "phonemic symbol string", for example, "sa" +
Convert to a symbol consisting of a continuous column of "ku" + "ra"
You. Further, the speech synthesis rule unit 204 sets the “phonological symbol
Of each voice unit from the voice unit storage unit 202 based on the
Load the data for The voice synthesizer 201 uses a waveform editing technique.
To produce a synthesized sound. Voice unit memory
The data loaded from the unit 202 is expressed in units of each CV.
Composed of waveform data used to generate synthesized sounds
Is done. The voice unit data used for waveform synthesis is as follows.
It consists of such a structure. The voiced part of the voice unit data is the actual voice
Of complex voiced parts using complex cepstrum analysis
Impulse and unit response waveform corresponding to one pitch
As one set, this set is required as one audio unit data
Audio data.
For unvoiced parts of data, cut off the waveform of unvoiced parts of real voice.
It consists of things that you take out and store. Therefore, the audio unit
If the data is in CV units, one voice unit CV
When the consonant part C is a voiceless consonant, cut out the unvoiced part
Waveform and multiple sets of impulse and unit response waveform
One voice unit data is constituted by
Impulse when consonant part C of unit CV is a voiced consonant
One voice by only multiple sets consisting of
Unit data is configured. The speech synthesis rule unit 204 includes a speech unit storage unit.
Text input of voice unit data loaded from 202
(Hereinafter, this data is referred to as synthesized waveform data.)
), And without text into the text,
It is possible to obtain a synthesized speech waveform whose power is read aloud. Further
The speech synthesis rule unit 204 performs the
The text input in the text analysis unit 203
Divide by length to detect breaks (i.e., poses)
Put out. As a result, as shown in FIG.
For example, as a text input, use the sentence
If you input "I got it,"
As such, the text input is “clean”,
"," "From Yamada", "I got it"
After that, between “Hana-no” and “Yamada-san”,
Is detected. Further, the speech synthesis rule unit 204 outputs
Rules and the basic accent of each word
To detect accents. Japanese phrase alone Accen
Is a unit of kana characters intuitively (hereinafter referred to as mora)
Can be expressed in two levels, high and low. this
Sometimes, depending on the content of the phrase, the accent position of the phrase
Can be distinguished. For example, ends, chopsticks, bridges
The gap is also a two-mora word that says “hashi”.
Type 0 without cents, accent position in first mora
There is a certain type 1 and a type 2 where the accent position is in the second mora
Can be classified. In this embodiment, the speech synthesis rule
As illustrated in FIG. 3 (C), the rule unit 204 performs the “text
The input clauses are called mora type 1, type 2, type 0, type 4
Classify and detect accents and poses on a phrase basis
You. Further, the speech synthesis rule section 204
Text input based on the
Generate a basic pitch pattern representing the intonation of the body. Japanese
In, the accent of the clause is intuitively at two levels
The actual intonation, while it can be represented, is illustrated in FIG.
As shown in the figure, the accent position gradually decreases.
There is a sign. Furthermore, in Japanese, clauses are
When one sentence comes, it follows the pose and goes to the next pose
Therefore, as illustrated in Fig. 3 (D), the intonation gradually decreases.
There is a feature to do. The speech synthesis rule section 204
Based on the characteristics of Japanese language
After generating the parameters to represent for each mora, a human utters
Mode so that the intonation changes smoothly as if
The parameters are set by interpolation between lines. From the above, the sound
The voice synthesis rule unit 204 determines the order according to the text input,
Combining the parameters of each mora and the interpolated parameters
(Hereinafter referred to as pitch pattern), read text input
Pitch pattern showing the intonation of the raised voice (Fig. 3 (F))
Get. The speech synthesizing unit 205 generates synthesized waveform data and
Performs waveform synthesis processing based on the pitch pattern
Generates a sound. This waveform synthesizing process is described below.
Perform by law. In the voiced part of the synthesized speech, the synthesized waveform
The impulse in the data is arranged based on the pitch pattern,
Unit response corresponding to each of the arranged impulses
The waveform is superimposed on each impulse. Also, the unvoiced part of the synthesized speech
Minute, the cut-out waveform in the synthesized waveform data is
The desired synthesized speech waveform is left as it is. This makes the pitch
Obtain a synthetic sound with varying intonation following the pattern change
be able to. Therefore, in the synthesized sound,
The pitch period of synthesized sound expands and contracts due to the use of impulses
Even so, there is almost no effect on sound source information,
Even when the switch pattern changes greatly
High quality that is similar to human voice without distortion of the envelope
A high quality arbitrary synthesized sound is obtained. In the above configuration, a predetermined input device
The input text input is sent to the
Is analyzed with reference to the dictionary of
This accent is detected. Word and phrase boundaries and bases
The detection result of the accent is obtained by the speech synthesis rule unit 204.
Processed according to the prescribed phonological rules described above, without inflection
Synthetic waveform data representing the speech read out of text input in the state
Data is generated. Plus words, phrase boundaries and basics
The result of accent detection is output by the speech synthesis rule
It is processed according to certain prosodic rules, and suppresses the entire text input.
A pitch pattern representing the fried fish is generated. Pitch pattern
Is output to the speech synthesizer 205 together with the synthesized waveform data.
Here, based on the pitch pattern and the synthesized waveform data,
Thus, a synthesized sound is generated. According to the above configuration, audio
In each of the audio unit data used for
Impulse and vocal tract as sound source information in voiced part
It has necessary sets of unit response waveforms as characteristics,
In the voice part, by having a cutout waveform of the real voice
Therefore, the distortion of the spectral envelope due to the change in pitch pattern
High quality synthesized sound that is close to human voice without causing distortion
Voices can be generated arbitrarily. In the first embodiment described above, the sound source information
Superimposed on the unit response waveform using impulse
The waveform is synthesized by
Is regarded as an ideal impulse,
Without superimposing them at pitch period intervals,
Position response waveforms so that they correspond to the pitch pattern.
Thus, a desired synthesized sound may be generated. Further
In the above-described embodiment, the voice unit storage unit 202
And holds voice unit data in CV units.
Is not only CV unit, but also another voice unit such as CVC unit
May hold the data. The phonetic symbol information 103 shown in FIG.
Sentence in the phoneme symbol string created by the speech synthesis rule unit 204 described above.
Recorded with knot accent and pause information added
Stored in device 102. The readout circuit 104 emits voice
The sound from the recording device 102 storing the phonetic symbol information 103
The voice pronunciation symbol information 103 is read. Voice synthesizer 105
Has a function equivalent to that of the speech synthesis unit 205 shown in FIG.
And the phonetic symbol information read out by the reading circuit 104
Based on 103, a synthesized speech output is obtained. According to the first embodiment, the recording device 102
The phonetic symbol information 103 to be remembered is the voice shown in FIG.
Since the sentence has been analyzed in advance by the synthesizer 201,
The storage capacity of the recording device 102 can be reduced. Ma
In the first embodiment, in the speech synthesizer 201,
The speech information analyzed in advance is the phonetic symbol information 10
3 is stored in the recording device 102,
There is no need to analyze the text and the playback time can be reduced. What
Note that, in the first embodiment, the reproduction of audio has been exemplified.
In this embodiment, not only sound but also sound can be widely reproduced.
Can be targeted. Sound is like sound
It is necessary to perform processing according to the sentence analysis and speech synthesis rules
I can't do it. The audio reproducing apparatus according to the first embodiment is applied to a vehicle.
Installed and played audio, or played audio and sound
Information as traffic guidance information for vehicles such as automobiles
Can be. In particular, like vehicles, spatial, price,
Quickly under conditions where power capacity is limited
If it is necessary to operate, the recording device 102
There is a real benefit in reducing the storage capacity and the playback time. The type of the recording device 102 is phonetic phonetic symbol information.
Information 103 storage capacity, quick playback time, and in some cases
Therefore, the determination is made in consideration of the possibility of correction. Quickness and
From this point of view, semiconductor memory is preferable, but capacity and
There may be price restrictions. For CD, MD, etc.
There is a slow playback speed, but in terms of price,
Less is. Considering the case of mounting on a vehicle,
Is affected by vibration, the recording device 102
A semiconductor memory that is not affected by vibration is preferable.
You. FIG. 4 shows a second embodiment of the signal reproducing apparatus according to the present invention.
Output the associated synthesized speech and text information as
1 shows an entire sound reproducing apparatus 401. Audio playback device 40
1 denotes a storage device 402, a first reading circuit 405, a sound
Voice synthesizer 406, second readout circuit 407, and
And a display device 408. The storage device 402 has a sound
Phonetic symbol information 404 and the text associated with it
Information 403 is stored. Text information 403
Is text information represented by ASCII code, etc.
You. The audio reproducing device 401 is a pair of audio phonetic symbol information.
Information 404 and text information 403 are reproduced at almost the same time.
You. Phonetic symbol information 404 and text information 4
03 may be the exact same sentence,
The text information 403 displayed on the display 408 is written
Yes, phonetic symbols output from the speech synthesizer 406
There can be a difference that the information 404 is a spoken language. Ma
Play different things for text display and voice playback.
Making it effective is also effective. For example, text display
The numbers are displayed and the audio gives only a comprehensive explanation. The first read circuit 405 is shown in FIG.
The same as the reading circuit 104,
The phonetic symbol information 404 is read. Second read time
The path 407 reads the text information 403 from the storage device 402.
Protrude. The speech synthesizer 406 is the speech synthesizer shown in FIG.
It is the same as the device 105. The display device 408 displays text information.
The information 403 is read by the second read circuit 407
Display the text information. For example, the display device 40
8 is a liquid crystal display, a cathode ray tube display, etc.
Can be used. Paired phonetic symbol information 404 and tele
To output the text information 403 at almost the same time,
Phonetic symbols paired with each piece of text information in device 402
Adds address information indicating the location where the information is recorded.
Can be kept. Also, paired phonetic phonetic information
404 and the text information 403 are linked in the storage device 402.
It is also possible to record in a continuous recording area.
In the case of, if no dead time is required for the address operation
There are advantages. FIG. 5 shows an embodiment of the present invention to which map information is newly added.
3 shows a third embodiment of the present invention. This playback device 5
01 is a phonetic transcription corresponding to the phonetic symbol information 404.
No. information 504, text corresponding to text information 403
Storage that stores map information 509 in addition to information 503
Device 502 corresponding to the first readout circuit 405.
The first reading circuit 505 outputs a voice from the storage device 502.
Read phonetic symbol information 504 and send it to speech synthesizer 406
Output to the corresponding speech synthesizer 506. Second read
The storage circuit 507 stores the text information 50 in the storage device 502.
3 and the map information 509 is read out and displayed on the display device 408.
Output to the corresponding display device 508. In the storage device 502
To record the map information 509, the signal value of each coordinate point
"1" or "0" and the position to display the text
You only need to know the information and the text content. This is map information 5
09 is black and white, but map information 509 is colored
For each coordinate point, three points, R, G, and B,
It is sufficient that the signal value is given. Stored in the storage device 502
Map information 509, phonetic symbol information 504, text
From the three pieces of information 503,
Output pictures and text as a map to guide the vehicle
This makes it desirable as a vehicle guidance information device.
You. FIG. 6 shows a fourth embodiment of the signal reproducing apparatus according to the present invention.
Is shown. The playback device 601 has a storage device 602 and a first reading device.
Readout circuit 605, speech synthesizer 606, second reading
A display circuit 607 and a display device 608. FIG.
Where the voice unit storage unit 202 is recorded with reference to
I did not mention it, but to improve the sound quality
However, since the amount of data for this will increase,
The playback device 601 shown in FIG.
To a low cost disk drive, etc.
Synthesize speech that is desirable to extract as needed
It takes time, but you can start production in advance
Putting it in practice does not cause inconvenience. Reproduction shown in FIG.
In the device 601, the phonetic transcription is stored in the storage device 602.
Number information 604, text information 603, map information 609
In addition, the speech synthesis unit 610 is stored, and the first
The read-out circuit 605 matches the phonetic symbol information 604 with the voice.
The speech element 610 is read out, and the speech is synthesized by the speech synthesizer 606.
Based on the phonetic symbol information 604 based on the synthesis segment 610,
It is configured to Map information 609 and text
The reproduction of the strike information 603 on the display device 608 is shown in FIG.
This is the same as the second embodiment described with reference to FIG. Voice synthesis unit
610 reduces the reproduction speed when stored in the semiconductor memory.
This is suitable for FIG. 7 shows a fifth embodiment of the signal reproducing apparatus according to the present invention.
Is shown. The playback device 701 shown in FIG.
2, a first readout circuit 705, a speech synthesizer 706,
A second reading circuit 707 and a display device 708 are provided.
The phonetic symbol information 704 and the speech synthesis are stored in the storage device 702.
Segment 710, text information 703 and map information 70
9 is stored. These configurations are again illustrated in FIG.
It is the same as the raw device 601. The playback device 70 shown in FIG.
1 further includes an internal control device 712 and a discriminating device.
711 and an input device 710 are provided. further
External control device 713 is connected to playback device 701
Have been. The input device 710 outputs the sound of the system.
User knows the text information output before the system
It is an input device for giving instructions.
A switch corresponding to S / N corresponds to this. Discriminator
The device 711 is used to output the audio of the system or text information.
Information output and the input signal from the input device
It is a discriminating device. A window on the time axis is used as shown in FIG.
Thus, a correspondence can be obtained. Input time window is on
The signal from the force device 710 is discriminated by the discrimination device 711 by the internal controller.
Troll device 712 or external control device 713
Can be passed only when the window is open, and at other times
Rejects. The input time window also has text information 703
Or the end of the presentation time of the output of the phonetic symbol information 704
After a while, for example, within about 10 seconds,
State is that the user's judgment and actions are text
This is because there is a possibility that it will be later than the presentation of the voice. Also
Of course, due to this, the input from the user is this excess time
The time difference above must be given to the next input
Become. In the vehicle guidance information device, the burden on the user
Not too fast is not necessary for the system to mitigate
This extra time within about 10 seconds is acceptable
You. The internal control device 712 is the discriminating device 7
11, input from input device 710 considered valid
The signal is received, the next operation is determined, and the storage device 702 is determined.
Control. For example, (1) The system uses the map information displayed to the user.
Hear or not to display the associated text information. (2) The user sends a yes response using the switch. (3) The discriminator 711 outputs a signal from the speech synthesizer 706
Then, know the voice utterance time width, create an input time window, and
A yes signal is output to the section control device 712. (4) The internal control device 712 controls the recording device 702
The corresponding text information 703 is output. External control
The troll device 713 is a storage device 7 in the playback device 701.
02 is a circuit that controls other than
Speed control of uttered voice of voice synthesizer 706
And so on. FIG. 9 shows a storage device 702 such as a CD or the like.
This shows the recording state when a disk is used. FIG. 9 (A)
901 indicates a storage area of the text information 703.
And reference numeral 902 appears at about the same time as the text information 703.
The storage area of the phonetic phonetic symbol information 704 to be input is shown. FIG.
FIG. 9B shows a storage method in comparison with FIG.
Storage area of text information 703 and phonetic symbol information 70
4 shows an example in which four storage areas are arranged at different positions. Text
Information 703 is stored in the areas indicated by reference numerals 903 and 904.
The phonetic symbol information 704 is represented by reference numerals 905 and 906.
It is stored in the indicated area. FIG. 10 shows a sixth embodiment of the signal reproducing apparatus of the present invention.
Here is an example. The signal reproducing apparatus shown in FIG.
An example of a signal reproducing device suitable as a guidance information device will be described. This
The vehicle guidance information device 1001 includes a storage device 1002,
A first readout circuit 1005, a speech synthesizer 1006,
Buffer memory 1016, second read circuit 100
7, display device 1008, input device 1010, discrimination device 1
011, the first controller 1012, the second controller
It has a trawl device 1013. These configurations themselves,
10 except that a buffer memory 1016 is provided.
7, a playback device 701 and an external control device.
The configuration is similar to that of the device 713, and its basic processing operation is shown in FIG.
7 is the same as the signal reproducing apparatus illustrated in FIG. The vehicle guidance information device 1001 further includes
The first control having a reset input device 1014
The device 1012 is provided with a selection information storage device 1015.
ing. The preset input device 1014 provides guidance information for the vehicle.
Route that the vehicle passes when using the alarm device 1000
Considering (course), select a group of playback audio in advance
This is a selection signal input device. Selection information storage device
Reference numeral 1015 denotes a preset input device 1014 and an input device.
The context from the input signal from the
Information storage device for selecting a set of reproduced voices according to
Holds phonetic symbol address information corresponding to voice input
I do. Second reading circuit 1007 and display device 1008
The buffer memory 1016 provided between the
A set of phonetic symbols that match the text is stored and
The sound is output via the synthesizing device 1006. With this configuration
Of the vehicle on which the vehicle guidance information device 1000 is mounted.
The driver outputs a sound output that takes into account the course of the vehicle.
From the device 1006 and from the display device 10
08. [0044] According to the present invention, a bit rate of about several hundred bps is obtained.
Get audio of contents related to the text to be displayed in the trait
Can be According to the present invention, the statement
Plays because chapter analysis and speech synthesis rule processing are performed.
Time can be reduced. Also has a sentence analysis unit and a dictionary
Low cost because you can configure an online system without having to
A signal reproducing device can be manufactured particularly. According to the present invention,
Voice output and input when the user inputs in response to voice output
We can attach correspondence with power.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a configuration diagram of a sound reproducing device as a first embodiment of a signal reproducing device of the present invention. FIG. 2 is a processing form diagram of a voice synthesizing apparatus that performs a text voice synthesizing process stored in a storage device of the voice reproducing apparatus shown in FIG. 1; FIG. 3 is a view for explaining generation of a basic pitch pattern in the speech synthesizer shown in FIG. 2; FIG. 4 is a configuration diagram of a sound reproducing device as a second embodiment of the signal reproducing device of the present invention. FIG. 5 is a configuration diagram of a sound reproducing device as a third embodiment of the signal reproducing device of the present invention. FIG. 6 is a configuration diagram of an audio reproduction device as a fourth embodiment of the signal reproduction device of the present invention. FIG. 7 is a configuration diagram of an audio reproducing apparatus as a fifth embodiment of the signal reproducing apparatus of the present invention. FIG. 8 is a diagram for explaining the operation of the discrimination device shown in FIG. FIG. 9 is a diagram illustrating a storage form of phonetic symbol information and text information when a disk device is used as a storage device in the audio reproduction device of the present invention. FIG. 10 is a configuration diagram of a vehicle guidance information device as a sixth embodiment of the signal reproduction device of the present invention. [Explanation of Reference Codes] 101, 401, 501, 601, 701... Playback devices 102, 402, 502, 602, 702, 1002
Recording devices 103, 404, 504, 604, 704, 1004
-Phonetic symbol information 403, 503, 603, 703, 1003-Text information 104, 405, 505, 605, 705, 1005-
First reading circuits 105, 406, 506, 606, 706, 1006
A voice synthesizing device 407, 507, 607, 707, 1007 second reading circuit 408, 508, 608, 708, 1008 display device 509, 609, 709, 1009 map information 610 Speech synthesis unit 201 Speech synthesis unit 202 Speech unit storage unit 203 Text analysis unit 204 Speech synthesis rule unit 205 Speech synthesis unit 710 Input device 711 · Discrimination device 712 ··· Internal control device 713 ··· External control device 1001 ··· Vehicle guidance information device 1014 ··· Preset input device 1015 ··· Selection information storage device 1016 ··· Buffer memory

Continuation of the front page (56) References JP-A-3-160500 (JP, A) JP-A-3-257485 (JP, A) JP-A-3-87799 (JP, A) JP-A-2-66598 (JP) JP-A-2-234199 (JP, A) JP-A-63-225299 (JP, A) JP-A-4-280300 (JP, A) JP-A-4-199196 (JP, A) 4-218871 (JP, A) JP-A-3-259200 (JP, A) JP-A-3-259196 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G10L 13/08 G10L 13/00

Claims (1)

(57) [Claims 1] Sound information including speech phonetic symbol information and a speech synthesis unit for speech synthesis that has been subjected to analysis processing for reproduction in advance, and the sound information corresponds to the sound information. Recording means for recording text information; selection information recording means for recording selection information of limited phonetic symbol information among the phonetic symbol information recorded in the recording means; and First reading means for reading the sound information from the recording means, sound synthesizing means for generating a sound reproduction signal from the read sound information, means for reproducing the generated sound reproduction signal, and second reading the text information And a display unit for displaying the text information.