JP2005024869A - Voice responder - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a voice responder which can easily be used by a user. <P>SOLUTION: The voice responder recognizes the uttering contents of the user and provides voice guidance which is beforehand set on the basis of the recognition result. In the device, uttering speed of the user is detected by uttering speed detecting means 12 and 13, and uttering contents of the responding voice guidance is changed in accordance with the detected uttering speed by uttering contents changing means 16 and 17. Even though the response time and uttering contents are the same, the uttering contents of the voice guidance is selectively changed and outputted in accordance with the uttering speed of the user, appropriate degree of proficiency is determined and proper measures are taken for the uttering contents of the voice guidance. Thus, the device can easily be used by the user. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a voice response device.
[0002]
[Prior art]
In general, this type of voice response device is used for, for example, reservation of a ticket using a telephone and reference / update of a database.
[0003]
Here, when this type of device is unfamiliar with the operation, there is not enough explanation of the operation of the device circulated by voice guidance, and it becomes difficult to understand what the user should do. Yes. On the other hand, when the user is familiar with the usage method, the content of the voice guidance is redundant, and it takes time for the user to complete the exchange.
[0004]
As a conventional technique considering such points, there is a voice response device that switches and sends voice guidance between an unfamiliar user and a skilled user (see, for example, Patent Documents 1 and 2). In these Patent Documents 1 and 2, the skill level of the user is determined based on the length of the reaction time from when the voice response device starts sending the guidance until the user responds by voice, and the guidance is changed according to the skill level. Like to do.
[0005]
Further, in Patent Document 3, when a user utters a word unnecessary for an operation or when there is a grudge, by reducing the skill level of the user, the user with a low skill level of the device Even when the reaction time is short, guidance for users who are unfamiliar with the operation can be sent out.
[0006]
[Patent Document 1]
JP-A-4-344930
[Patent Document 2]
Japanese Patent Laid-Open No. 10-20884
[Patent Document 3]
JP 2001-331196 A
[0007]
[Problems to be solved by the invention]
However, in these conventional voice response devices as described above, the reaction time is the same because the utterance speed such as a slow utterance or a fast utterance is not considered. In addition, when the content of the utterance is correct, there is a problem that it is determined that the skill level is the same between the user unfamiliar with the operation and the user accustomed to the operation.
[0008]
This is more likely to occur, for example, when the utterance content is simple, such as a number sequence or “Yes” or “No”. For example, in the case of both the person who speaks the sequence “0471” for a short time and the person who speaks for a long time, If the content is correct and the reaction time is the same, it is determined that the skill level is the same. Therefore, the voice response device is not always easy to use for the user.
[0009]
In addition, the voice guidance that is selectively output when the skill level is low is originally “Are you sure?”, Whereas “Are you sure? Speak“ Yes ”if you want.” As described above, the description of the operation will be described in more detail as compared with the normal time, and the output voice quality, for example, the volume and the speech rate will not be changed. Therefore, it tends to remain difficult to understand after all for users who are unfamiliar with the operation. In this respect as well, the voice response device is not always easy for the user to use.
[0010]
Recently, the number of users who use voice response devices via mobile phones has increased, and even if the user has difficulty in hearing voice guidance, such as when there is a lot of ambient noise, No change processing is performed on the guidance output. Therefore, when using a mobile phone or the like, the voice response device is not always easy to use for the user.
[0011]
An object of the present invention is to provide a voice response device that is easy for a user to use.
[0012]
[Means for Solving the Problems]
The present invention relates to a voice response device for recognizing a user's utterance content and providing a preset voice guidance based on the recognition result, an utterance speed detecting means for detecting the utterance speed of the user, and a detected utterance Utterance content changing means for changing the utterance content of the voice guidance that responds according to the speed, even if the reaction time and the utterance content are the same, the skill level is appropriately determined according to the utterance speed of the user By selectively changing and outputting the utterance content of the voice guidance according to the result, the voice response device is easy to use for the user.
[0013]
DETAILED DESCRIPTION OF THE INVENTION
A first embodiment of the present invention will be described with reference to FIGS.
The voice response device of the present embodiment shows an application example to a ticket reservation server using a telephone or the like, for example. The device itself is basically a computer with a built-in microcomputer and is installed or Various processes are executed according to the computer program stored and saved, including the processes described below.
[0014]
First, a hardware configuration example of the server 1 serving as the voice response device according to the present embodiment will be described with reference to a schematic block diagram shown in FIG. The server 1 has, as its schematic configuration, a CPU 2 that executes various arithmetic processes and centrally controls each unit, a ROM 3 that stores fixed data, a RAM 4 that stores variable data in a rewritable manner, a display device, a speaker, and the like Output device 5, various input devices 6, storage device 7, communication interface 8, and the like are connected via a system bus 9.
[0015]
For example, a hard disk drive is used as the storage device 7. In the storage device 7, various programs are installed in addition to the OS, and various files, tables, and the like related to ticket reservation are stored and saved. The OS, programs, and various files installed in the storage device 7 are all or partially copied to the RAM 4 when the program of the server 1 is started, so that the access speed is increased.
[0016]
In addition, the server 1 can be accessed by voice by the user via the communication interface 8, a telephone, or the like.
[0017]
FIG. 2 shows a functional block diagram of the voice response device (server 1) of the present embodiment using such a computer. In the voice response device according to the present embodiment, first, a voice input unit 11 through which a user inputs voice, a voice analysis unit 12 that analyzes and detects a speech rate of voice input to the voice input unit 11, and a speech A voice analysis storage unit 13 using a ROM 3 in which conditions for determining the speed are set is provided. Here, the speech analysis unit 12 and the speech analysis storage unit 13 realize the function of the speech rate detection means. Further, a voice recognition unit 14 that recognizes input voice and outputs a voice recognition result, and a ROM 3 or a storage device in which words and phrases necessary for the operation of the automatic voice response device are set in advance for the user to receive a service. 7 is used. Further, the dialogue processing storage unit 16 using the ROM 3 or the storage device 7 in which the content of the dialogue processing performed by the voice response device (server 1) is set in advance, and the voice recognition result and the setting contents of the dialogue processing storage unit 16 Dialogue processing unit 17 for managing speech content of speech guidance, guidance storage unit 18 using ROM 3 or storage device 7 in which speech data of speech content of speech guidance is set, and speech guidance speech from guidance storage unit 18 A guidance output unit 19 that acquires audio data of contents and outputs the audio data, and an audio output unit 20 that outputs the audio data from a speaker in the output device 5 are provided. Here, the dialogue processing unit 17 and the dialogue processing storage unit 16 realize the function of the utterance content changing means.
[0018]
In such a configuration, when a user inputs a voice via the communication interface 8 or the like, the voice information input to the voice input unit 11 is sent to the voice analysis unit 12. Here, the processing flow of the voice analysis unit 12 executed by the CPU 2 in accordance with the program will be described with reference to a schematic flowchart shown in FIG.
[0019]
First, in step S1, the input sound is divided into frames, and an acoustic power level is obtained for each divided frame. Is the utterance part for all frames by comparing each frame with the reference value previously determined for the intermediate value between the maximum sound power level and the minimum sound power level in all frames? Identify whether it is the utterance part. Then, the utterance time of the input voice is obtained by adding the time of the frame that is the utterance part. The obtained result is stored in the internal memory (RAM 4).
[0020]
In step S2, the input voice is transmitted to the voice recognition unit 14, and a recognition result is awaited. If the speech recognition is normally performed based on the recognition result obtained from the speech recognition unit 14 (Y in S2), the process proceeds to step S3. If it cannot be recognized (N in S2), the speech rate determination result is set to "normal speech rate" and stored in the internal memory (RAM 4), and the process proceeds to step S4.
[0021]
In step S3, the number of words is counted from the speech recognition result obtained from the speech recognition unit 14. For example, if the voice recognition result is “good”, the number of words is counted as it is, such as 3 words, and if it is “Chiketto”, the number of words is counted as it is. Then, the number of utterance words per second is calculated as the “utterance speed” from the utterance time obtained in step S1. Then, the calculated speech rate is compared with the speech rate condition preset in the speech analysis storage unit 13, the speech rate of the input speech is determined, and the determination result is stored in the internal memory (RAM 4). The process proceeds to step S4.
[0022]
4 is an example of setting the speech rate condition preset in the speech analysis storage unit 13. Here, as the speech rate condition, “slow utterance” is used when the number of speech words per second is 3.8 or less. If it exceeds 8 and is 4.2 or less, “normal utterance speed” is set, and if it exceeds 4.2, “fast utterance” is set.
[0023]
At this time, if the obtained utterance speed is 3.6, for example, the utterance speed of the input voice is determined as “slow utterance” and set in the internal memory (RAM 4).
[0024]
In step S4, the speech recognition result and the speech speed determination result set in the internal memory (RAM 4) are transmitted to the dialog processing unit 17 and the process is terminated.
[0025]
The dialogue processing unit 17 selects the field of voice guidance to be output according to the contents of the dialogue processing preset in the dialogue processing storage unit 16 from the voice recognition result. For the selected voice guidance field, the speech guidance utterance content list preset in the dialogue processing storage unit 16 is referred to, and the speech guidance having the utterance content corresponding to the utterance speed of the user is selected. Then, the selected voice guidance is transmitted to the guidance output unit 19.
[0026]
FIG. 5 shows a table example of an utterance content list of voice guidance based on the utterance speed of a user in the field of “ticket reservation”, which is preset in the dialogue processing storage unit 16 (ROM 3 or the like). In the example shown in the figure, for example, when the speech rate is “slow utterance”, “Reservation is OK? If yes, please say“ Yes ”, if it is wrong, please say“ No ””, “Normal speech rate” The voice guidance utterance content is set to "Ticket reservations received. How many days are reservations?", And "Fast utterances" is "How many days are reservations?" Here, for example, if the result of speech recognition is “ticket reservation” and the judgment result of the user's utterance speed is “normal utterance speed”, “ticket reservation accepted. Select the voice guidance for the utterance content, and if it is “slow utterance”, the voice content of the utterance content will say “Yes, if you like ticket reservation? Guidance will be selected.
[0027]
The guidance output unit 19 acquires the voice data of the corresponding voice guidance from the guidance storage unit 18 from the utterance content of the selected voice guidance and transmits it to the voice output unit 20.
[0028]
The sound output unit 20 outputs the transmitted sound data as output sound.
[0029]
With the above processing, the utterance content of the voice guidance is changed and output according to the utterance speed. Therefore, it is usually a voice guidance of the utterance content “You have booked a ticket. How many days will it be?”, But for users who are unfamiliar and have a slow utterance speed, For example, it is a voice guidance of polite utterance contents such as “Ticket reservation? If yes, please say“ Yes ”, if wrong, please say“ No ”, which is easy for the user to use. Conversely, for users who are accustomed and have a fast utterance speed, the voice guidance for the minimum utterance content is not redundant and the user is easy to use. Become. Particularly, since the user who is unfamiliar and the user who is accustomed are discriminated by paying attention to the utterance speed, even when the reaction time is the same and the utterance content is correct, the skill level can be discriminated appropriately.
[0030]
A second embodiment of the present invention will be described with reference to FIGS. The same parts as those shown in the first embodiment are denoted by the same reference numerals, and description thereof is also omitted (the same applies to the subsequent embodiments in order).
[0031]
In the present embodiment, with respect to the configuration (functional block) shown in FIG. 2, on the output side of the guidance output unit 19, data conversion processing means by the voice conversion unit 21, the voice conversion storage unit 22, and the speech speed conversion unit 23. Is added. Here, the voice conversion unit 21 performs a change instruction for changing the speech speed (speaking speed) of the voice guidance input from the guidance output unit 19 in accordance with the conversion condition and a volume conversion process. The ROM 3 or the storage device 7 In the voice conversion storage unit 22 using the voice conversion conditions for performing the voice data conversion process, the voice conversion conditions are set in advance. Further, the speech speed conversion unit 23 executes speech speed conversion processing for voice guidance. The speech conversion unit 21, the speech conversion storage unit 22, and the speech speed conversion unit 23 change the speech speed and speech volume of the speech guidance, that is, the speech quality of the speech guidance.
[0032]
In the present embodiment, one kind of voice guidance utterance content set in the guidance storage unit 18 is prepared for each voice guidance field. For example, in the case of a ticket reservation field, a normal “ticket” The content of the utterance is stored, "Reservation accepted.
[0033]
The speech guidance utterance content list preset in the dialogue processing storage unit 16 sets the speech content of the same voice guidance for the speech speed.
[0034]
In such a configuration, when a user inputs a voice, the voice information input to the voice input unit 11 is sent to the voice analysis unit 12. The speech analysis unit 12 performs the same processing as in the above-described embodiment, and transmits the speech recognition result and the speech rate determination result set in the internal memory to the dialogue processing unit 17. The dialogue processing unit 17 selects the field of voice guidance to be output according to the contents of the dialogue processing preset in the dialogue processing storage unit 16 from the voice recognition result. In addition, the voice guidance is selected with reference to the utterance content list of the voice guidance preset in the dialogue processing storage unit 16 for the selected voice guidance field. Then, the selected voice guidance and utterance speed determination result are transmitted to the guidance output unit 19.
[0035]
The guidance output unit 19 acquires voice data corresponding to the utterance content of the voice guidance from the guidance storage unit 18 from the selected voice guidance and transmits it to the voice conversion unit 21. The received speech speed determination result is also transmitted to the voice conversion unit 21.
[0036]
The voice conversion unit 21 performs a conversion process on the received voice data according to the speech speed determination result, and then transmits the converted voice data to the voice output unit 20.
[0037]
FIG. 7 is a schematic flowchart showing a processing flow of the voice conversion unit 21. In step S11, the voice conversion unit 21 first stores the voice data input from the guidance output unit 19 in the internal memory. In the next step S <b> 12, the speech rate determination result given to the speech conversion unit 21 is compared with speech conversion conditions preset in the speech conversion storage unit 22.
[0038]
FIG. 8 shows voice conversion conditions preset as a table in the voice conversion storage unit 22 using the ROM 3 or the like, that is, conversion conditions of the volume and speech rate of the output voice guidance. When the user's speaking speed is “slow speaking”, the volume of the output voice guidance is 1.2 times, the speaking speed is 0.9 times, and when the “normal speaking speed”, the volume and speaking speed are not changed. In the case of “fast speech”, the volume is not changed and the speech speed is set to 1.2 times. Here, for example, if the determination result of the utterance speed is “slow utterance”, the volume is 1.2 times higher than the volume conversion condition.
[0039]
In step S12, the determination result of the speech rate given to the voice conversion unit 21 is compared with the volume conversion condition of the output voice guidance set in the voice conversion storage unit 22. If the conversion condition is satisfied, the process goes to step S13. If the condition is not met, the process proceeds to step S14. In step S13, the volume of the audio data stored in the internal memory is converted, and after updating the audio in the internal memory, the process proceeds to step S14.
[0040]
In step S14, the determination result of the speech rate given to the speech conversion unit 21 is compared with the speech rate conversion condition of the output speech guidance set in the speech conversion storage unit 22, and if the condition is satisfied, the process goes to step S15. If the condition does not apply, the process proceeds to step S16.
[0041]
Here, for example, if the determination result of the utterance speed is “slow utterance”, the speech speed of the output voice guidance is 0.9 times based on the conversion condition of the utterance speed of the output voice guidance in FIG. Since the speaking speed is changed, the process proceeds to step S15.
[0042]
In step S15, the voice data in the internal memory (RAM 4) is transmitted to the speech speed conversion unit 23. After the voice data in the internal memory (RAM 4) is updated with the received voice data, the process proceeds to step S16. For example, if the utterance speed is to be converted to 1.2 times, voice data and an instruction to increase the utterance speed of the voice data to 1.2 times are given to the speech speed conversion unit 23, and as a result, the utterance speed is increased. Will receive the voice data that has increased by a factor of 1.2.
[0043]
In step S16, the audio data in the internal memory (RAM 4) is transmitted to the audio output unit 20 and the process ends. The audio output unit 20 outputs the transmitted audio data from the speaker as output audio.
[0044]
Through the above processing, the voice data conversion process is performed to change the volume of the voice guidance to be output and the speech speed in accordance with the speech speed of the input voice. When the speech speed is slow, the volume is set to 1. • Voice guidance with an utterance speed that is 0.9 times slower at 2 times is output, and when the utterance speed is fast, voice guidance with an utterance speed increased to 1.2 times without changing the volume is output. The
[0045]
With the above processing, for example, even if the voice guidance utterance content is the same as “Ticket reservation accepted. How many days will it be?” Since the voice guidance with 1.2 times the utterance speed and 0.9 times slower is output, the voice guidance is easy to hear and can be heard without hurry, which makes it easy for the user to use. On the contrary, in the case of a user who is used and has a high utterance speed, the utterance speed is increased and the voice guidance finish time is shortened. Therefore, the user is not redundant in time and is easy to use for the user. In the case of the present embodiment as well, since the user who is unfamiliar and the user who is accustomed are discriminated by paying attention to the speaking speed, the skill level is properly discriminated even when the reaction time is the same and the utterance content is correct can do.
[0046]
In this embodiment, as an example of changing the voice quality of voice guidance, the example of volume and speed has been described. However, only one of them may be used, or the voice pitch, voice pitch, etc. may be changed as the voice quality. You may make it make it (it is the same also in subsequent embodiment).
[0047]
A third embodiment of the present invention will be described with reference to FIGS. In the present embodiment, in the configuration (functional block) shown in FIG. 6, the speech analysis unit 12 and the speech analysis storage unit 13 realize the function of the speech rate detection means and detect the noise situation related to the user's speech. The function of the noise detection means is realized. In other words, the voice analysis unit 12 is provided with a function of analyzing the noise state of the voice, and the voice analysis storage unit 13 using the ROM 3 or the RAM 4 is preset with a condition for determining the noise state.
[0048]
In such a configuration, when a user inputs a voice, the voice information input to the voice input unit 11 is sent to the voice analysis unit 12.
[0049]
FIG. 9 is a schematic flowchart showing the flow of processing of the voice analysis unit 12 of the present embodiment. Note that the processing control relating to the speech rate shown in FIG. 3 is omitted. In step S21, the voice analysis unit 12 divides the input voice in units of frames and obtains an acoustic power level for each divided frame. An intermediate value between the maximum sound power level and the minimum sound power level among all frames is used as a reference value, and comparison processing is performed for each frame, thereby distinguishing whether all frames are utterance portions or not utterance portions. Then, the ratio of the average acoustic power level of the frame that is the utterance part and the average acoustic power level of the frame that is not the utterance part is obtained, and the result is compared with the condition of the noise situation preset in the speech analysis storage unit 13; Determine ambient noise conditions. Then, the determined result is stored in the internal memory (RAM 4), and the process proceeds to step S22.
[0050]
Here, FIG. 10 is an example of the condition of the noise situation preset as a table in the voice analysis storage unit 13 using the ROM 3 or the like. Here, the condition of the noise situation is 0.7 or less when “low noise”, If it exceeds 0.7, “high noise” is set in two stages. At this time, if the average sound power level of the section in which the obtained utterance is performed is 80 and the average sound power level of the section in which the utterance is not performed is 70, the ratio is 0.75, and the above condition is satisfied. The noise situation is determined to be “noisy” and set in the internal memory (RAM 4).
[0051]
In step S22, the input voice is transmitted to the voice recognition unit 14, and a recognition result is awaited. Then, the recognition result obtained from the voice recognition unit 14 and the determination result of the noise situation set in the internal memory are transmitted to the dialogue processing unit 17 and the process is terminated.
[0052]
The dialogue processing unit 17 selects a field of voice guidance to be output according to the content of the dialogue processing set in advance in the dialogue processing storage unit 16 from the voice recognition result. For the selected voice guidance field, the speech guidance utterance content list preset in the dialogue processing storage unit 16 is referred to, and the speech guidance having the utterance content corresponding to the utterance speed of the user is selected. Then, the selected voice guidance, speech rate determination result, and noise status determination result are transmitted to the guidance output unit 19.
[0053]
The content list of the voice guidance is set as shown in FIG. 11, for example. That is, guidance items (speech contents of voice guidance) are classified according to the user's utterance speed. The guidance output unit 19 acquires voice data corresponding to the utterance content of the voice guidance from the guidance storage unit 18 from the selected voice guidance and transmits it to the voice conversion unit 21. The received speech rate determination result and noise status determination result are also transmitted to the voice conversion unit 21.
[0054]
The voice conversion unit 21 converts the received voice data according to the speech speed determination result and the noise situation determination result, and then transmits the converted voice data to the voice output unit 20.
[0055]
The voice conversion unit 21 performs voice processing according to the flowchart of FIG. In step S11, the voice conversion unit 21 stores the voice data input from the guidance output unit 19 in the internal memory. In the next step S12, the speech rate determination result and noise state determination result given to the speech conversion unit 21 and the output speech guidance volume and speech rate conversion conditions preset in the speech conversion storage unit 22 are obtained. Compare.
[0056]
FIG. 12 shows the conversion conditions for the volume and speech rate of the output voice guidance preset as a table in the voice conversion storage unit 22 using the ROM 3 or the like. Even if the utterance speed of the user is the same, the volume and the speech speed of the voice guidance are set so as to be switched between two levels according to the size of the noise situation. In particular, when the noise state is “noisy”, the volume is set to be high and the speech speed is set to be low.
[0057]
In step S12, the determination result of the speech speed and the noise situation is compared with the sound volume conversion condition of the output voice guidance. If the conversion condition is satisfied, the process proceeds to step S13. If the conversion condition is not satisfied, the process proceeds to step S14.
[0058]
In step S13, the volume of the audio data stored in the internal memory is converted, and after updating the audio in the internal memory, the process proceeds to step S14.
[0059]
In step S14, the speech rate determination result, the noise state determination result, and the speech rate conversion condition of the output voice guidance are compared. If the condition is satisfied, the process proceeds to step S15. If the condition is not satisfied, the process proceeds to step S16. move on.
[0060]
Therefore, under such setting conditions, for example, the speech recognition result is “ticket reservation”, the speech speed determination result is “normal speech speed”, and the noise situation determination result is “low noise”. If there is, the voice guidance of the utterance content “Your ticket reservation is accepted. How many days is reserved?” Is output at the normal level for both the volume and the speech speed, and the judgment result of the noise situation is “noisy” For example, the voice guidance of the utterance content “You received a ticket reservation. How many days will it be reserved?” Is increased to 1.2 times the volume and output at a speed that is 0.9 times slower. Become. The same applies to other utterance speeds.
[0061]
Through the above processing, the utterance content of the voice guidance is selected according to the utterance speed and the noise situation, and the volume and the speech speed are changed and output as the utterance voice quality. Therefore, according to the present embodiment, even if the reaction time and utterance content are the same, the utterance content of the voice guidance is selectively changed according to the utterance speed of the user and the noise situation around the user. In addition to outputting, voice data is processed to change the voice quality when outputting voice guidance, and in situations where there is a lot of noise, the voice guidance becomes easier to hear by slowing the utterance speed or increasing the volume, It is easy for users to use regardless of whether they are familiar or unfamiliar. This is particularly effective in an environment where a mobile phone or the like is used.
[0062]
A fourth embodiment of the present invention will be described with reference to FIG. In the present embodiment, in the configuration (functional block) shown in FIG. 6, as in the third embodiment, the speech analysis unit 12 and the speech analysis storage unit 13 realize the function of the speech rate detection unit, The function of the noise detection means for detecting the noise situation related to the user's utterance is realized. That is, a function for analyzing the noise situation of the voice is added to the voice analysis unit 12, and conditions for determining the noise situation are preset in the voice analysis storage unit 13 using the ROM 3 or the like. In the present embodiment, only one type of speech guidance utterance content is prepared for each guidance field. For example, in the case of the field of ticket reservation, the usual “ticket reservation is accepted. It is assumed that the utterance content “is it a reservation?” Is stored.
[0063]
FIG. 13 shows the conversion conditions of the volume and speech rate of the output voice guidance preset as a table in the voice conversion storage unit 22 using the ROM 3 or the like. The volume and the voice speed of the voice guidance are set so as to be switched between two levels according to the user's voice speed and noise level. In particular, when the noise state is “noisy”, the volume is set to be high and the speech speed is set to be low.
[0064]
The process control in the present embodiment is the same as in the third embodiment, except that the speech guidance speech content is not selectively switched according to the speaker speed.
[0065]
Therefore, under such setting conditions, for example, if the speech recognition result is “ticket reservation”, the speech speed determination result is “slow utterance”, and the noise situation determination result is “low noise” , "Voice reservations are accepted. How many days is the reservation?" The voice guidance of the utterance content is output at a volume of 1.2 times and a speech speed of 0.9 times. ”, The voice guidance of the same utterance content“ The ticket reservation was accepted. How many days will it be reserved? ”Was further increased to 1.3 times the volume and slowed to 0.8 times the speaking speed. Will be output. The same applies to other utterance speeds.
[0066]
Therefore, according to the present embodiment, even when the reaction time and the content of the utterance are the same, the voice quality of the utterance is changed when outputting the voice guidance according to the utterance speed of the user and the noise situation around the user. Voice data is processed to make changes, and in a noisy situation, the voice guidance becomes easier to hear by lowering the utterance speed and increasing the volume, regardless of whether you are used or unfamiliar It will be easy to use. This is particularly effective in an environment where a mobile phone or the like is used.
[0067]
A fifth embodiment of the present invention will be described with reference to FIGS. In the present embodiment, in the configuration (functional block) shown in FIG. 6, the voice analysis unit 12 and the voice analysis storage unit 13 realize the function of a noise detection unit that detects a noise situation related to the user's utterance, The function of the speech rate detection means is deleted. That is, in the present embodiment, the skill level discrimination function based on the speech rate is omitted, the voice analysis unit 12 is added with a function of analyzing the noise status of the voice, and the voice analysis storage unit 13 has the noise status. A condition for discriminating is set in advance.
[0068]
In such a configuration, when a user inputs a voice, the voice information input to the voice input unit 11 is sent to the voice analysis unit 12. The processing flow of the voice analysis unit 12 of the present embodiment is the same as that described in the flowchart of FIG.
[0069]
Here, FIG. 15 is an example of the condition of the noise situation preset as a table in the voice analysis storage unit 13. Here, the condition of the noise situation is “small noise” when 0.6 or less, and exceeds 0.6. If it is 0.7 or less, it is set in three stages, “medium”, and if it exceeds 0.7, “large noise”. At this time, if the average sound power level of the section in which the obtained utterance is performed is 80 and the average sound power level of the section in which the utterance is not performed is 70, the ratio is 0.75, and the above condition is satisfied. The noise situation is determined as “noisy” and set in the internal memory.
[0070]
In the present embodiment, only one type of speech guidance utterance content is prepared for each guidance field. For example, in the case of the field of ticket reservation, the usual “ticket reservation is accepted. Utterance contents are stored.
[0071]
FIG. 15 shows the conversion conditions for the volume and speech rate of the output voice guidance preset as a table in the voice conversion storage unit 22 using the ROM 3 or the like.
[0072]
It is set so that the volume and the voice speed of the voice guidance can be switched in three stages according to the noise level. In particular, when the noise state is “noisy”, the volume is set to be higher and the speech speed is set to be slower.
[0073]
Therefore, under such setting conditions, for example, if the result of speech recognition is “ticket reservation” and the judgment result of the noise situation is “low noise”, “ticket reservation accepted. If the result of the noise situation is “noisy”, the voice guidance of the utterance content is output at a normal level and the speech speed is 1.2 times faster than normal. The voice guidance of the utterance content, “Ticket reservation accepted. How many days is reserved?” Is increased to 1.2 times the volume and output at a slower speed of 0.9 times.
[0074]
Through the above processing, the volume and the speech speed are changed and output as the speech quality of the speech guidance according to the noise situation of the environment where the user is speaking. Therefore, according to the present embodiment, even if the reaction time and the content of the utterance are the same, the voice data of the voice data is changed in order to change the utterance voice quality when outputting the voice guidance according to the noise situation around the user. In the situation where the noise is high, the voice guidance is easy to hear by slowing down the utterance speed or increasing the volume, and it is easy for the user to use. This is particularly effective in an environment where a mobile phone or the like is used.
[0075]
【The invention's effect】
According to the present invention, it is possible to provide a voice response device that is easy for a user to use.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a first embodiment of the present invention.
FIG. 2 is a functional block diagram thereof.
FIG. 3 is a schematic flowchart showing processing contents in the voice analysis unit;
FIG. 4 is an explanatory diagram showing an example of a condition for determining an utterance speed.
FIG. 5 is an explanatory diagram showing an example of utterance contents according to the utterance speed.
FIG. 6 is a functional block diagram showing a second embodiment of the present invention.
FIG. 7 is a schematic flowchart showing processing contents in the voice conversion unit;
FIG. 8 is an explanatory diagram showing a setting example of a volume and a speech speed according to the speech speed.
FIG. 9 is a schematic flowchart illustrating processing contents in a speech analysis unit according to a third embodiment of this invention.
FIG. 10 is an explanatory diagram illustrating an example of a noise condition determination condition;
FIG. 11 is an explanatory diagram showing an example of utterance contents according to the utterance speed and noise status.
FIG. 12 is an explanatory diagram illustrating a setting example of a volume and a speech speed according to a speech speed and a noise situation.
FIG. 13 is an explanatory diagram illustrating a setting example of a sound volume and a speech speed according to an utterance speed and a noise state according to the fourth embodiment of this invention.
FIG. 14 is an explanatory diagram illustrating an example of a noise condition determination condition according to the fifth embodiment of this invention;
FIG. 15 is an explanatory diagram illustrating a setting example of a volume and a speech speed according to a noise situation.
[Explanation of symbols]
12, 13 Speech rate detection means, noise detection means
16, 17 Utterance content change means
21-23 Data conversion processing means

Claims (5)

In a voice response device that recognizes a user's utterance content and provides a preset voice guidance based on the recognition result,
Utterance speed detecting means for detecting the utterance speed of the user;
Utterance content changing means for changing the utterance content of the voice guidance to respond according to the detected utterance speed;
A voice response device comprising: In a voice response device that recognizes a user's utterance content and provides a preset voice guidance based on the recognition result,
Utterance speed detecting means for detecting the utterance speed of the user;
Data conversion processing means for performing voice data conversion processing for changing the voice quality of the voice guidance that responds according to the detected speech speed;
A voice response device comprising: Noise detecting means for detecting a noise situation related to the user's utterance;
Data conversion processing means for performing voice data conversion processing for changing the voice quality of the voice guidance that responds according to the detected noise situation;
The voice response device according to claim 1, further comprising: Noise detecting means for detecting a noise situation related to the user's utterance;
The data conversion processing means performs voice data conversion processing for changing the voice quality of the voice guidance that responds according to the detected speech speed and noise status.
The voice response device according to claim 2. In a voice response device that recognizes a user's utterance content and provides a preset voice guidance based on the recognition result,
Noise detecting means for detecting a noise situation related to the user's utterance;
Data conversion processing means for performing voice data conversion processing for changing the voice quality of the voice guidance that responds according to the detected noise situation;
A voice response device comprising:

Images