Sound processing method, program, and sound processing system

Abstract

In a sound processing method, event information relating to an event is acquired from an information source device or an information source service (S1); a step (S4) of determining a character string to cause the speech device to sound on the basis of the acquired event information; dividing the determined character string into one or more sub-character strings (S5); generating a sound signal according to the character string (S7); a correction step (S6, S8) for correcting the generated sound signal by performing a filtering process in which a first filter corresponding to the characteristics of the consonants is applied for each sub-string; and outputting the corrected sound signal (S11).

Description

Technical Field

[0001] The present disclosure relates to a sound processing method and the like for processing a human voice emitted from a speaker. Background Art

[0002] For example, Patent Document 1 discloses a voice response automatic adjustment method for improving the intelligibility of a human voice received in a wireless receiver by automatically adjusting the voice response according to the ambient noise level. In this voice response automatic adjustment method, if there is a large ambient noise, the relative gain of the high sound frequency is increased at the expense of the low frequency response.

[0003] Prior art literature

[0004] Patent Literature

[0005] Patent Document 1: Japanese Patent Publication No. 2000-508487 Summary of the invention

[0006] Problems to be solved by the invention

[0007] The present disclosure provides a sound processing method and the like that allows a user to easily hear a human voice without relying on the performance of a speaker included in a speech device.

[0008] Means used to solve problems

[0009] In a sound processing method related to a technical solution of the present disclosure, event information related to an event is obtained from an information source device or an information source service; based on the obtained event information, a character string that causes a speaking device to speak is determined; the determined character string is divided into one or more sub-character strings; a sound signal is generated based on the character string; for the generated sound signal, filtering processing is performed by applying a first filter corresponding to the characteristics of the consonant to each of the sub-character strings, thereby modifying the sound signal; and the modified sound signal is output.

[0010] A program according to one technical solution of the present disclosure causes one or more processors to execute the above-mentioned sound processing method.

[0011] A sound processing system related to a technical solution of the present disclosure comprises: an input interface for obtaining event information related to an event from an information source device or an information source service; a signal processing circuit for modifying a sound signal; an output interface for outputting the modified sound signal; the signal processing circuit for determining a character string for causing a speaking device to speak based on the obtained event information; dividing the determined character string into one or more sub-character strings; generating a sound signal based on the character string; and for the generated sound signal, performing filtering processing of applying a first filter corresponding to the characteristics of a consonant to each of the sub-character strings, thereby modifying the sound signal.

[0012] Effects of the Invention

[0013] According to the voice processing method and the like of the present disclosure, there is an advantage that a user can easily hear a human voice without depending on the performance of a speaker included in a speaking device. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a diagram showing the frequency characteristics of a human voice when a speech device is caused to output a human voice.

[0015] Figure 2 1 is a diagram showing the frequency characteristics of a human voice when a speech device outputs a human voice after filtering a sound signal.

[0016] Figure 3 This is a diagram showing the frequency characteristics of the driving sound output by the electrical device.

[0017] Figure 4 This is a diagram showing the results of performing DRT by causing the robot cleaner to output evaluation sounds in a noisy environment.

[0018] Figure 5 This is a diagram showing the results of the characteristics of each consonant when DRT is performed by causing the robot cleaner to output an evaluation sound in a noisy environment.

[0019] Figure 6 This is a diagram showing the results of the characteristics of each consonant when DRT is performed by causing a pet camera to output an evaluation sound in a noisy environment.

[0020] Figure 7 This is a diagram showing an example of a spectrogram obtained from the sound waveform of the evaluation sound for each characteristic of the consonant.

[0021] Figure 8 This is a diagram showing an example of a spectrogram obtained from the sound waveform of the evaluation sound for each characteristic of the consonant.

[0022] Fig. 9 This is a block diagram showing an example of the overall configuration of a sound processing system according to the embodiment.

[0023] Fig.10 This is a diagram showing the correspondence between consonants and their features.

[0024] Fig.11 It is a diagram showing the correspondence between the features of consonants and the frequency domain.

[0025] Fig.12 This is a flowchart showing an example of the operation of the sound processing system according to the embodiment. DETAILED DESCRIPTION

[0026] [1. Knowledge underlying the present disclosure]

[0027] First, the inventor's ideas will be described below.

[0028] Conventionally, there is a technology for making a home appliance (speaking device) with a sound input / output function speak by instructing the content of speech and the timing of speech. Here, "sound" refers to vibrations of air or the like that can be perceived by humans through at least hearing. This technology is used to make the content of an event that occurs in a home appliance such as a washing machine be spoken from a speaking device with a sound input / output function, thereby notifying a user who is next to the home appliance of the content of the event that occurred. For example, an event may include a certain error occurring in the home appliance or the completion of an action performed by the home appliance.

[0029] Here, if the speaking device is a device that is assumed to mainly output human voices, such as a television, the performance of the speaker provided by the speaking device will be relatively high, and the user can easily hear the human voice output by the speaking device, that is, the intelligibility of the human voice is relatively high. On the other hand, if the speaking device is a device that is assumed to mainly output system sounds other than human voices such as beeps, such as a robot cleaner, the performance of the speaker provided by the speaking device will be relatively low, and it may be difficult for the user to hear the human voice output by the speaking device, that is, the intelligibility of the human voice is relatively low.

[0030] Therefore, in view of the above problems, the inventors have studied a technology that allows a user to easily hear a human voice, that is, a technology that makes the intelligibility of a human voice relatively high, without relying on the performance of a speaker included in a speaking device.

[0031] [1-1. Frequency characteristics of human voice output by a speech device]

[0032] First, the inventors studied improving the intelligibility of human voice output by a speech device by performing filtering processing on a sound signal converted into sound waves and output by the speech device. The filtering processing described here is processing to amplify the power (sound pressure level) in a specific frequency band of the sound signal.

[0033] Figure 1 : is a graph showing the frequency characteristics of a human voice when a speech device is made to output a human voice. Figure 2 : is a graph showing the frequency characteristics of a human voice when a speech device outputs a human voice after filtering a sound signal. Figure 1 and Figure 2 In all graphs, the vertical axis represents the power of human voice, and the horizontal axis represents the frequency.

[0034] Figure 1(a) is a diagram showing the frequency characteristics of human voice output by a robot cleaner as a speaking device, Figure 2 (a) is a diagram showing the frequency characteristics of human voice output by the robot cleaner when filtering is performed. Figure 1 (b) is a diagram showing the frequency characteristics of human voice output by a pet camera as a talking device, Figure 2 (b) is a diagram showing the frequency characteristics of human voice output by the pet camera when filtering is performed. In addition, both the robot cleaner and the pet camera are devices that are assumed to mainly output system sounds other than human voices.

[0035] like Figure 1 As shown in (a) of FIG. 1 , the power of the human voice output by the robot cleaner is lower in the low frequency range of 0 to 1 kHz and the high frequency range above 3 kHz than in other frequency ranges (see the circular marks in the figure). Therefore, filtering processing is performed on the sound signal to amplify the power in the above low frequency range and high frequency range. Figure 2 As shown in (a), the power of the human voice output by the robot cleaner increases in both the low-frequency domain and the high-frequency domain, and it is found that filtering processing can help improve the intelligibility of the human voice.

[0036] like Figure 1 As shown in (b) of FIG. 1 , the power of the human voice output by the pet camera in the low frequency range of 0 to 1 kHz and the high frequency range above 4 kHz is lower than that in other frequency ranges (see the circular marks in the figure). Therefore, as described above, filtering processing is performed on the sound signal to amplify the power in the low frequency range and the high frequency range. However, as Figure 2 As shown in (b), the power of the human voice output by the pet camera does not increase in the above-mentioned low-frequency domain and high-frequency domain, and it is recognized that the filtering process cannot help improve the intelligibility of the human voice.

[0037] As described above, the inventors have come to the following understanding: when filtering is performed according to the frequency characteristics of a person's voice output by a speech device, the filtering process can help improve the intelligibility of the person's voice depending on the type of speech device, but there are cases where the filtering process cannot help improve the intelligibility of the person's voice. Hereinafter, this filtering process is sometimes referred to as "filtering process corresponding to the speech device".

[0038] [1-2. DRT test with Japanese voice]

[0039] Next, the inventors conducted a DRT (Diagnostic Rhyme Test), a two-choice voice intelligibility test, under Japanese voice by making the speech device output human voice in a noisy environment. Here, the noisy environment is an environment where the driving sound (noise) is output by the driving of electrical devices located around the speech device.

[0040] Figure 3 This is a graph showing the frequency characteristics of the driving sound output by the electrical device. Figure 3 In the figure, the vertical axis represents the power of driving the sound, and the horizontal axis represents the frequency. Figure 3 (a) shows the frequency characteristics of the driving sound output by the sweeper. Figure 3 (b) shows the frequency characteristics of the driving sound output by the robot cleaner. Figure 3 (c) shows the frequency characteristics of the driving sound output by the dishwasher. Figure 3 (d) shows the frequency characteristics of the driving sound output by the washing machine.

[0041] DRT is an intelligibility test method that makes subjects listen to one of the candidate word pairs that differ only in the first phoneme and select one of the word pairs. In DRT, consonants are divided into 6 features, 10 pairs of word pairs are prepared for each feature, and a total of 120 words are evaluated. In DRT, the intelligibility of human voice is expressed by (number of correct answers - number of incorrect answers) / total number of evaluation sounds.

[0042] Here, consonants are divided into six characteristics: voice, nasality, sustainability, sibilation, graveness, and compactness.

[0043] The vocality corresponds to "vocalic-nonvocalic" in the characteristic classification of English phonemes based on JFH (Jacobson, Fant, Halle) (hereinafter referred to as "JFH classification"), and is a classification of voiced and unvoiced. Voiced is a sound accompanied by the vibration of the vocal cords, such as "ざい(zai)", and unvoiced is a sound not accompanied by the vibration of the vocal cords, such as "さい(sai)".

[0044] Nasal is equivalent to "nasal-oral" in the JFH classification and is a classification of nasal sounds. For example, nasal sounds are sounds in which the sound energy is not emitted from the mouth but radiates through the nose, such as "man" (man), or sounds in which the sound energy is emitted from the mouth but radiates through the nose, such as "ban" (ban).

[0045] In the JFH classification, continuity is equivalent to "continuant-interrupted", and is a classification of continuous sounds and other sounds (plosives or affricates). For example, continuous sounds are sounds that are not restricted by the narrowness of the vocal tract until the sound of / h / flows out, such as "hashi". For example, non-continuous sounds are plosive sounds such as "kashi".

[0046] Irregularity corresponds to "strident-mellow" in the JFH classification and is a classification related to the irregularity of the waveform. An example of a sound with irregularity is a sound like "tyaku", and an example of a sound without irregularity is a sound like "kaku".

[0047] The muteness corresponds to "grave-acute" in the JFH classification, and corresponds to mute and acute. Mute is, for example, a sound like "pai", and acute is, for example, a sound like "tai".

[0048] Concentration is equivalent to "compact-diffuse" in the JFH classification, and is a classification of whether the energy on the spectrum is concentrated on one formant frequency or dispersed. The former is, for example, a sound like "yaku", and the latter is a sound like "waku".

[0049] Figure 4 : is a graph showing the results of performing DRT by causing the robot cleaner to output evaluation sound in a noisy environment. Figure 4 In the above figure, the vertical axis represents the intelligibility of human voice (Speech Intelligibility, language clarity), and the horizontal axis represents the type of noise source. Figure 4 In FIG. 1 , the solid-line-shaded bar graphs represent the results when the filtering process corresponding to the speech device is not performed, and the dot-shaded bar graphs represent the results when the filtering process is performed. Figure 4 (a) shows the result when the robot cleaner outputs the evaluation sound in a female voice. Figure 4 (b) shows the result when the robot cleaner outputs the evaluation sound in a male voice.

[0050] like Figure 4As shown in the figure, when any of the electrical devices, including the cleaning machine, dishwasher, robot cleaning machine, and washing machine, is the noise source, the improvement in the intelligibility of the human voice is confirmed by performing the above-mentioned filtering process. In addition, it is confirmed that the intelligibility of the human voice is slightly higher when the robot cleaning machine outputs the evaluation sound with a female voice than when the robot cleaning machine outputs the evaluation sound with a male voice. However, when any electrical device is the noise source, the intelligibility of the human voice is relatively low, being less than 0.4.

[0051] As described above, the inventors have found that simply performing filtering processing according to a speaking device cannot contribute to sufficient improvement of the intelligibility of a person's voice in a noisy environment.

[0052] Here, the inventors have conducted a more detailed study on the above-mentioned DRT, and specifically, have conducted a study on the intelligibility of human voices for the features of each consonant in the DRT. Figure 5 This is a diagram showing the results of the characteristics of each consonant when DRT is performed by causing the robot cleaner to output an evaluation sound in a noisy environment. Figure 6 This is a diagram showing the results of the characteristics of each consonant when DRT is performed by causing a pet camera to output an evaluation sound in a noisy environment.

[0053] exist Figure 5 and Figure 6 In each, the vertical axis represents the intelligibility of human voice, and the horizontal axis represents the type of consonant feature. Figure 5 and Figure 6 In each of the above, the solid-line-shaded bar graphs represent the results when the filtering process corresponding to the speech device is not performed, and the dot-shaded bar graphs represent the results when the filtering process is performed. Figure 5 (a) and Figure 6 (a) shows the result when the noise source is a robot cleaner. Figure 5 (b) and Figure 6 (b) shows the result when the noise source is a washing machine.

[0054] like Figure 5 and Figure 6 As shown in FIG. 1 , it can be seen that, due to the characteristics of the consonants, even when the above-mentioned filtering process is performed, there are cases where the subjects cannot hear the evaluation sound clearly. Figure 5As shown in FIG. 1 , when the robot cleaner outputs evaluation sounds, the intelligibility of the evaluation sounds corresponding to the characteristics of audibility and irregularity is relatively high, but the intelligibility of the evaluation sounds corresponding to the characteristics of other consonants is relatively low. In particular, the intelligibility of the evaluation sounds corresponding to the characteristics of nasality and persistence is very low, and the subjects can hardly hear the evaluation sounds at all. In addition, for example, Figure 6 As shown, when the pet camera outputs an evaluation sound and the washing machine is the noise source, the intelligibility of the evaluation sound corresponding to the audibility is relatively high, but the intelligibility of the evaluation sound corresponding to the characteristics of other consonants is very low and the subjects can hardly hear the evaluation sound at all.

[0055] Therefore, the inventors focused on the characteristic frequency characteristics of each consonant. Figure 7 and Figure 8 Each of them is a diagram showing an example of a spectrogram obtained from the sound waveform of the evaluation sound based on the characteristics of each consonant. Figure 7 and Figure 8 In each of the above diagrams, the upper area represents the sound waveform, and the lower area represents the spectrogram. The spectrogram here is a diagram showing the frequency spectrum of the evaluation sound over time.

[0056] Figure 7 (a) shows a spectrogram obtained from the sound waveform of the evaluation sound "ざい(zai)" corresponding to the acoustic property. Figure 7 (b) shows a spectrogram obtained based on the sound waveform of the evaluation sound "さい(sai)" corresponding to the audibility. Figure 7 (c) shows a spectrogram obtained from the sound waveform of the evaluation sound "man" corresponding to the nasal sound. Figure 7 (d) shows a spectrogram obtained based on the sound waveform of the evaluation sound "ban" corresponding to the nasality. Figure 7 (e) shows a spectrogram obtained from the sound waveform corresponding to the continuous evaluation sound "hashi". Figure 7 (f) shows a spectrogram obtained based on the sound waveform corresponding to the continuous evaluation sound "かし (kasi)".

[0057] Figure 8 (a) shows a spectrogram obtained from the sound waveform of the evaluation sound "ちゃく(tyaku)" corresponding to irregularity, Figure 8 (b) shows a spectrogram obtained from the sound waveform of the evaluation sound "kaku" corresponding to irregularity. Figure 8 (c) shows a spectrogram obtained from the sound waveform of the evaluation sound "ぱい(pai)" corresponding to the sound suppression performance. Figure 8(d) shows a spectrogram obtained from the sound waveform of the evaluation sound "tai" corresponding to the sound suppression performance.

[0058] Figure 8 (e) shows a spectrogram obtained from the sound waveform of the evaluation sound "yaku" corresponding to the concentration. Figure 8 (f) shows a spectrogram obtained from the sound waveform of the evaluation sound "waku" corresponding to the concentration.

[0059] like Figure 7 and Figure 8 As shown in , the frequency spectrum of the first phoneme of a word is different according to the characteristics of each consonant. Figure 7 (a) and Figure 7 As shown by the rectangular frame in (b), if we focus on the spectrogram of the evaluation sound corresponding to the acoustic property, in "ざ(za)" and "さ(sa)" corresponding to the first phoneme of the word, the former contains a frequency component of 0 to 1 kHz, while the latter does not. Figure 8 (a) and Figure 8 As shown in the rectangular box in (b) of FIG. 1 , if we focus on the spectrogram of the evaluation sound corresponding to irregularity, the former contains more frequency components of 2 to 6 kHz, while the latter contains almost none. In the characteristics of other consonants, as shown in FIG. Figure 7 and Figure 8 As indicated by the arrow or rectangular box in the figure, the frequency spectrum in the first phoneme of the word is different.

[0060] Here, in Figure 5 (a) and Figure 5 In (b), if we focus on the results of the evaluation sound corresponding to the vocality, the intelligibility of the human voice is greatly improved when the filtering process corresponding to the speech device is performed, compared with the case where the filtering process is not performed. This is considered to be because the effective frequency range for hearing the first phoneme of the evaluation sound corresponding to the vocality is 0 to 1 kHz, and the filtering process that amplifies the power in the low frequency range of 0 to 1 kHz is emphasized.

[0061] As described above, the inventors have found that by emphasizing the frequency domain corresponding to the characteristics of the consonant in the voice signal, it is easier for humans to hear the first phoneme of a word, and as a result, the intelligibility of human voice is improved.

[0062] In view of the above, the inventors created the present disclosure.

[0063] Hereinafter, the embodiments are described in detail with reference to the accompanying drawings. In addition, the embodiments described below all represent inclusive or specific examples. The numerical values, shapes, materials, constituent elements, configuration positions of constituent elements and connection forms, steps, the order of steps, etc. shown in the following embodiments are examples and are not intended to limit the present disclosure. In addition, regarding the constituent elements of the following embodiments that are not described in the independent claims, they are described as arbitrary constituent elements.

[0064] In addition, each figure is a schematic diagram, and does not necessarily illustrate the exact figure. In addition, in each figure, the same reference numerals are given to substantially the same structure, and the repeated description may be omitted or simplified.

[0065] (Implementation Method)

[0066] [2. Structure]

[0067] [2-1. Overall structure]

[0068] First, use Fig. 9 The overall configuration of a sound processing system including the embodiment will be described. Fig. 9 1 is a block diagram showing the overall structure of a sound processing system including the embodiment. The sound processing system 10 is a system for causing a human voice representing the content of an event to be output from a speech device 2 when event information about the event is obtained from an information source device 3 or an information source service 4. In the embodiment, the human voice is a human voice based on Japanese. In addition, in the embodiment, the sound processing system 10 is implemented by a server 1. The server 1 can communicate with the speech device 2, the information source device 3, and the information source service 4 respectively via a network N1 such as the Internet. In addition, the server 1 can also communicate with a part or all of the speech device 2, the information source device 3, and the information source service 4 via a LAN (Local Area Network).

[0069] In the embodiment, the server 1 (sound processing system 10) causes one speaker 2 to output the voice of a person indicating the content of an event, but the server 1 may cause a plurality of speaker devices 2 to respectively output the voice of a person indicating the content of an event. In addition, the server 1 may cause one or more of the plurality of speaker devices 2 to output the voice of a person indicating the content of an event. In addition, the server 1 may cause the content of the event to be spoken to be different for each speaker device 2. For example, the server 1 may cause one of the two speaker devices 2 to output the voice of a person indicating the content of an event related to the information source device 3, and the other speaker device 2 to output the voice of a person indicating the content of an event related to another information source device 3.

[0070] The speaker device 2 is a device capable of notifying the user of the content of an event by outputting a human voice indicating the content of the event generated in the information source device 3 or the information source service 4. The notification by the speaker device 2 may also be performed by, for example, displaying a character string or an image on a display of the device.

[0071] The speech device 2 is, for example, a device installed in a facility where the user lives, and is a device having the sound output function as described above. In an embodiment, the speech device 2 is a household appliance. Specifically, the speech device 2 may include, for example, a smart speaker, a television, a lighting fixture, a pet camera, a headphone mother machine, a headphone daughter machine, an air conditioner, or a robot cleaner. In addition, the speech device 2 may also be, for example, a portable information device held by a user, such as a portable television, a smart phone, a tablet terminal, or a laptop personal computer.

[0072] The information source device 3 is a device that is a transmission source of the information spoken from the speech device 2. In the embodiment, the information source device 3 is a household appliance. Specifically, the information source device 3 is, for example, an air conditioner, a washing machine, a cleaner, a robot cleaner, a dishwasher, a refrigerator, a rice cooker, or a microwave oven. The event generated in the information source device 3 may include, for example, the start or end of an action of the information source device 3, the occurrence of an error in the information source device 3, or the maintenance of the information source device 3. Fig. 9 In the figure, there is one information source device 3, but there may be multiple information source devices.

[0073] The information source service 4 is a service that is a transmission source of the information spoken from the speech device 2, and is, for example, a service provided to users by a server operated by a service provider. The information source service 4 is, for example, a transportation service, a weather forecast service, a schedule management service, or a traffic information service. The events generated in the information source service 4 may include, for example, the start or end of the service provided by the information source service 4, or the occurrence of an error in the information source service 4. Fig. 9 In the example, the information source service 4 is one, but it may be multiple.

[0074] [2-2. Server structure]

[0075] Next, the structure of the server 1 is described in detail. Fig. 9 As shown, the server 1 includes a communication interface (hereinafter referred to as “communication I / F (Interface)”) 11 , a processor 12 , a memory 13 , and a storage device 14 .

[0076] The communication I / F 11 is, for example, a wireless communication interface, and communicates with the information source device 3 or the information source service 4 via the network N1 based on a wireless communication standard such as Wi-Fi (registered trademark), thereby receiving signals sent from the information source device 3 and the information source service 4. In addition, the communication I / F 11 communicates with the speaker device 2 via the network N1 based on a wireless communication standard such as Wi-Fi (registered trademark), sends signals to the speaker device 2, and receives signals sent from the speaker device 2.

[0077] The communication I / F 11 has the functions of an input interface (hereinafter referred to as "input I / F") 11A and an output interface (hereinafter referred to as "output I / F") 11B. The input I / F 11A receives a signal sent from the information source device 3 or the information source service 4, and obtains event information related to the event from the information source device 3 or the information source service 4.

[0078] In the embodiment, the input I / F 11A further acquires sound collection information obtained by collecting sounds around the speaker 2. The sound collection information is, for example, information related to sound data generated by collecting sounds by a microphone 25 (described later) provided in the speaker 2. The sound around the speaker 2 is noise that makes it difficult for the user to hear the voice of a person who expresses the content of an event when the speaker 2 outputs the voice of the person. The input I / F 11A acquires the sound collection information by receiving the sound data transmitted from the speaker 2 as the sound collection information.

[0079] The output I / F 11B outputs the sound signal corrected by the processor 12 by sending a signal to the speaker 2. In addition, the output I / F 11B outputs an instruction signal instructing the speaker 2 to collect surrounding sounds by sending a signal to the speaker 2.

[0080] The processor 12 is, for example, a CPU (Central Processing Unit) or a DSP (Digital Signal Processor), and performs information processing related to the transmission and reception of signals using the communication I / F 11, and information processing for generating and correcting sound signals based on event information obtained by the communication I / F 11. The above-mentioned processing related to the transmission and reception of signals and the information processing for generating and correcting sound signals are all realized by the processor 12 executing a computer program stored in the memory 13. The processor 12 is an example of a signal processing circuit of the sound processing system 10.

[0081] The memory 13 is a storage device that stores various information necessary for the processor 12 to perform information processing and a computer program executed by the processor 12. The memory 13 is implemented by, for example, a semiconductor memory.

[0082] The storage device 14 is a device that stores a database that is referred to when the processor 12 performs information processing for generating and correcting an audio signal. The storage device 14 is implemented by, for example, a semiconductor memory such as a hard disk or an SSD (Solid State Drive).

[0083] [2-3. Structure of the speaking device]

[0084] Next, the structure of the speaking device 2 will be described in detail.

[0085] like Fig. 9 As shown, the speaker 2 includes a communication I / F 21, a processor 22, a memory 23, a speaker 24, and a microphone 25. Depending on the type of the speaker 2, the speaker 2 may not include the microphone 25. In the following description, it is assumed that the speaker 2 includes the microphone 25 unless otherwise specified.

[0086] The communication I / F 21 is, for example, a wireless communication interface, and communicates with the server 1 via the network N1 based on a wireless communication standard such as Wi-Fi (registered trademark), thereby receiving a signal transmitted from the server 1 and transmitting a signal to the server 1 .

[0087] The processor 22 is, for example, a CPU or a DSP, and performs information processing related to the transmission and reception of signals using the communication I / F 21, information processing for causing the microphone 25 to collect sounds around the speech device 2 based on an instruction signal received by the communication I / F 21, and information processing for causing the speaker 24 to output a person's voice based on a sound signal received by the communication I / F 21. The above-mentioned information processing related to the transmission and reception of signals, information processing for outputting a person's voice, and information processing for collecting sounds around the speech device 2 are all realized by the processor 22 executing a computer program stored in the memory 23.

[0088] The memory 23 is a storage device that stores various information necessary for the processor 22 to perform information processing and a computer program executed by the processor 22. The memory 23 is implemented by, for example, a semiconductor memory.

[0089] The speaker 24 reproduces a human voice based on the voice signal received by the communication I / F 21. In the embodiment, the speaker 24 converts the voice signal into a human voice and outputs the converted human voice.

[0090] The microphone 25 collects the sounds around the speech device 2 and generates sound data. In an embodiment, the microphone 25 does not always collect the sounds around the speech device 2, but only collects the sounds around the speech device 2 when receiving an instruction to pick up sound from the server 1 (the sound processing system 10). The sound data generated by the microphone 25 is sent to the server 1 as sound pickup information via the communication I / F 21.

[0091] [2-4. Generation and Correction of Sound Signals]

[0092] Next, a detailed description will be given of the information processing for generating and correcting sound signals performed by the processor 12 of the server 1 (the sound processing system 10).

[0093] First, if the communication I / F 11 (input I / F 11A) acquires event information, the processor 12 determines the speech device 2 that outputs the human voice representing the content of the event. When there is only one speech device 2, the processor 12 determines to make this speech device 2 output the human voice. In addition, when there are multiple speech devices 2, the processor 12 determines to make a preset speech device 2 among the multiple speech devices 2 output the human voice. At this time, the speech device 2 that outputs the human voice is not limited to one, and may also be multiple.

[0094] Next, the processor 12 outputs an instruction signal to the determined speech device 2 via the communication I / F 11 (output I / F 11B) so that the speech device 2 collects the surrounding sounds. Thereby, the processor 12 acquires the sound pickup information from the determined speech device 2 via the communication I / F 11 (input I / F 11A). In addition, when the determined speech device 2 does not have the microphone 25, the processor 12 does not acquire the sound pickup information from this speech device 2.

[0095] Next, the processor 12 determines a string representing the content of the event based on the acquired event information. For example, when acquiring event information indicating the end of the washing operation from the washing machine as the information source device 3, the processor 12 determines a string such as "洗濯機での洗濯が終わりました (The washing in the washing machine has ended)". In an embodiment, the processor 12 automatically generates a string based on the event information through an appropriate automatic generation algorithm.

[0096] In addition, the processor 12 may also determine the string by, for example, referring to a database stored in the storage device 14 and reading out the string corresponding to the acquired event information. In this case, data associating the content of the event with the string corresponding to the event is stored in the database in advance.

[0097] Next, the processor 12 divides the determined character string into one or more sub-character strings by using an appropriate algorithm. In an embodiment, the processor 12 divides the determined character string into one or more sub-character strings based on syllables. Here, a syllable is a segmentation unit for dividing continuous speech sounds, and is a collection of the auditory sense of human voice. For example, a syllable includes a consonant, a vowel, a consonant+vowel, a vowel+consonant, or a consonant+vowel+consonant.

[0098] In the implementation, as an example, the processor 12 divides the determined character string into one or more sub-character strings according to the following rules. First, the processor 12 divides the character string into one or more sub-character strings for each consonant and each vowel, and on this basis, long sounds, short sounds, and dial sounds are considered as one sub-character string together with the immediately preceding consonant. In addition, the processor 12 may also combine a vowel with the immediately preceding consonant and consider it as one sub-character string. For example, when the character string contains "washing machine (せんたくき)(sentakuki)", the processor 12 divides the character string into four sub-character strings, namely "せん(sen)", "た(ta)", "く(ku)", and "き(ki)".

[0099] Next, the processor 12 determines the first filter to be applied for each sub-string. Here, the first filter is a filter corresponding to the characteristics of the consonant, and is a filter for amplifying and emphasizing the power in the frequency domain corresponding to the characteristics of the consonant. In addition, the processor 12 does not apply the first filter to the sub-string consisting of only vowels.

[0100] Fig.10 This is a diagram showing the correspondence between consonants and their features. Fig.10 (a) is a table that lists consonants corresponding to the characteristics of each consonant (voice, nasality, persistence, irregularity, suppression, and concentration). For example, the consonants having the characteristic of voice are the ka row, sa row, ta row, ga row, za row, and da row. Fig.10 (b) is a table listing the features of consonants corresponding to each consonant. For example, the consonants in the ka row have four consonant features: voicing, sustention, sibilation, and compactness.

[0101] Fig.11This is a diagram showing the correspondence between the characteristics of consonants and the frequency domain. For example, in a consonant having the characteristic of voicedness, the effective frequency domain for hearing the first phoneme of a word is 0 to 1 kHz. In addition, in a consonant having the characteristic of nasality, the effective frequency domain for hearing the first phoneme of a word is 1 to 4 kHz.

[0102] Fig.10 and Fig.11 The data shown is stored in a database stored in the storage device 14. Then, the processor 12 determines the first filter to be applied for each sub-character string by referring to the data stored in the database.

[0103] For example, in the case of four sub-strings "せん(sen)", "た(ta)", "く(ku)", and "き(ki)", the processor 12 determines the sub-string "せん(sen)" based on the fact that "せ(se)" included in "せん(sen)" is in the sa row. Fig.10 The corresponding relationship shown in the figure determines the voice, persistence and irregularity as the characteristics of the consonant corresponding to the sa line. In addition, the processor 12 synthesizes the filter that amplifies the power of the frequency domain corresponding to the voice, the filter that amplifies the power of the frequency domain corresponding to the persistence, and the filter that amplifies the power of the frequency domain corresponding to the irregularity and determines them as the first filter.

[0104] In addition, regarding the sub-character string "た(ta)", the processor 12 calculates the sub-character string "た(ta)" based on the fact that "た(ta)" is a ta line. Fig.10 The corresponding relationship shown in the figure determines the voice, persistence, irregularity, and muteness as the characteristics of the consonants corresponding to the ta row. In addition, the processor 12 synthesizes the filter that amplifies the power of the frequency domain corresponding to the voice, the filter that amplifies the power of the frequency domain corresponding to the persistence, the filter that amplifies the power of the frequency domain corresponding to the irregularity, and the filter that amplifies the power of the frequency domain corresponding to the muteness and determines them as the first filter.

[0105] In addition, the processor 12 generates sub-character strings such as "く(ku)" and "き(ki)" based on the fact that "く(ku)" and "き(ki)" are both ka (ka) lines. Fig.10 The corresponding relationship shown in FIG. 1 determines the voice, persistence, irregularity, and concentration as the characteristics of the consonant corresponding to the ka line. Furthermore, the processor 12 synthesizes and determines as the first filter the filter that amplifies the power of the frequency domain corresponding to the voice, the filter that amplifies the power of the frequency domain corresponding to the persistence, the filter that amplifies the power of the frequency domain corresponding to the irregularity, and the filter that amplifies the power of the frequency domain corresponding to the concentration.

[0106] Furthermore, when there are a plurality of consonant features for each sub-character string, the processor 12 may determine each of the filters corresponding to the features of the plurality of consonants as the first filter without synthesizing the filters corresponding to the features of the plurality of consonants.

[0107] Next, the processor 12 generates a sound signal based on the determined character string by using an algorithm for generating an appropriate machine sound. In the embodiment, the processor 12 generates a sound signal based on a female voice based on the knowledge that the intelligibility of a female voice is slightly higher than that of a male voice in [1. The knowledge that is the basis of the present disclosure]. Alternatively, the processor 12 may generate a sound signal based on a male voice.

[0108] Next, the processor 12 performs a process of correcting the generated sound signal. In the embodiment, the processor 12 performs a first filtering process of applying a first filter to the generated sound signal, a second filtering process of applying a second filter, and a third filtering process of applying a third filter. In addition, these three filtering processes can be performed in the above order or in other orders.

[0109] In the embodiment, before executing the first filtering process, the processor 12 causes the memory 13 to store the position (time) corresponding to each sub-string in the generated sound signal. For example, when the sound signal is generated based on the string "せんたくき(sentakuki)", the processor 12 causes the memory 13 to store that the sub-string "せん(sen)" corresponds to the time of 0 to 0.7 seconds of the sound signal, the sub-string "た(ta)" corresponds to the time of 0.7 seconds to 1 second of the sound signal, the sub-string "く(ku)" corresponds to the time of 1 to 1.3 seconds of the sound signal, and the sub-string "き(ki)" corresponds to the time of 1.3 seconds to 1.6 seconds of the sound signal.

[0110] Furthermore, in the first filtering process, the processor 12 applies the first filter determined for each sub-string to the position (time) corresponding to each sub-string stored in the memory 13. For example, when the first filter is applied to the sub-string "せん(sen)", the processor 12 applies the first filter to the time from 0 to 0.7 seconds of the audio signal.

[0111] In addition, the processor 12 is not limited to the above-mentioned processing. For example, the processor 12 may generate a sound signal and apply the first filter for each sub-string, and then generate a sound signal corresponding to the determined string by connecting the sound signals corresponding to all the sub-strings. In other words, the processor 12 may generate a sound signal corresponding to the determined string by generating and connecting the sound signals for each sub-string. In this case, the processor 12 may apply the corresponding first filter to each sound signal generated in sub-string units and perform correction. However, when the sound signal is generated and connected for each sub-string, it may feel awkward when a person hears it, so the former method is preferred.

[0112] Here, the second filter is a filter corresponding to the type of the speech device 2. Specifically, the second filter is a filter that amplifies and emphasizes the power of a frequency domain with relatively low power based on the frequency characteristics of the human voice output by the speech device 2. For example, when the speech device 2 is a robot cleaner, the power of the human voice output by the robot cleaner is lower in the low frequency domain of 0 to 1 kHz and the high frequency domain above 3 kHz compared to other frequency domains (refer to Figure 1 In this case, the processor 12 determines the filter that amplifies the power of the frequency range of 0 to 1 kHz and the frequency range of 3 kHz or more as the second filter.

[0113] In addition, the third filter is a filter corresponding to the sound reception information. Specifically, the third filter is a filter that amplifies and emphasizes the power of the frequency domain with relatively high power based on the frequency characteristics of the sound around the speaking device 2 obtained from the sound reception information. For example, it is assumed that a dishwasher is operating around the speaking device 2. In this case, the processor 12 analyzes the frequency characteristics of the sound data included in the sound reception information and determines that the power is relatively high in the frequency domain of 0 to 500 Hz (refer to Figure 3 (c)). Therefore, in this case, the processor 12 determines the filter that amplifies the power in the frequency domain of 0 to 500 Hz as the third filter.

[0114] The frequency characteristics of the surrounding sounds of the speaker 2 may be calculated by the processor 22 of the speaker 2 when the speaker 2 collects the surrounding sounds, or may be calculated by the processor 12 of the server 1 that has acquired the collected sound information.

[0115] Then, the processor 12 transmits (outputs) the voice signal corrected by executing the first filtering process, the second filtering process, and the third filtering process to the determined speaker 2 via the communication I / F 11 (output I / F 11B). As a result, the speaker 2 as the target obtains the corrected voice signal via the communication I / F 21, and reproduces the human voice based on the corrected voice signal from the speaker 24.

[0116] [3.Action]

[0117] Below, use Fig.12 An operation of the server 1 (sound processing system 10 ) according to the embodiment, that is, an example of a sound processing method will be described. Fig.12 This is a flowchart showing an example of the operation of the sound processing system 10 according to the embodiment. In the following, it is assumed that an event occurs in the information source device 3 or the information source service 4, and the event information is sent from the event source to the server 1 via the network N1. In addition, it is assumed that the speech device 2 has a microphone 25 and can provide the sound collection information to the server 1.

[0118] First, the processor 12 acquires event information via the communication I / F 11 (input I / F 11A) (S1). Then, the processor 12 determines the speaking device 2 that outputs a human voice indicating the content of the event (S2).

[0119] Next, the processor 12 outputs an instruction signal to the determined speaker 2 to collect surrounding sounds via the communication I / F 11 (output I / F 11B). Thus, the processor 12 obtains sound collection information from the determined speaker 2 via the communication I / F 11 (input I / F 11A) (S3).

[0120] Next, the processor 12 determines a character string representing the content of the event based on the acquired event information by using an appropriate automatic generation algorithm (S4). Furthermore, the processor 12 divides the determined character string into one or more sub-character strings by using an appropriate algorithm (S5). Here, the processor 12 divides the determined character string into one or more sub-character strings based on syllables.

[0121] Next, the processor 12 determines the first filter to be applied for each sub-string (S6). Here, the processor 12 refers to the database stored in the storage device 14. Fig.10 and Fig.11 The data shown determines the first filter to be applied for each sub-character string.

[0122] Next, the processor 12 generates a voice signal based on the determined character string using an algorithm for generating an appropriate machine voice (S7). Here, the processor 12 generates a voice signal based on a female voice.

[0123] Next, the processor 12 performs a first filtering process (S8) of applying a first filter to each sub-string on the generated sound signal. As described above, in the embodiment, the processor 12 causes the memory 13 to store the position (time) corresponding to each sub-string for the generated sound signal. And, in the first filtering process, the processor 12 applies the first filter determined for each sub-string to the position (time) corresponding to each sub-string stored in the memory 13. In addition, the processor 12 performs a second filtering process (S9) of applying a second filter on the generated sound signal. In addition, the processor 12 performs a third filtering process (S10) of applying a third filter on the generated sound signal. In addition, the order of executing steps S8, S9, and S10 is not limited to this order, and may be another order.

[0124] Next, the processor 12 transmits (outputs) the modified audio signal to the determined speaker 2 via the communication I / F 11 (output I / F 11B) (S11). As a result, the speaker 2 as the target obtains the modified audio signal via the communication I / F 21, and reproduces the human voice based on the modified audio signal from the speaker 24.

[0125] [4. Effects, etc.]

[0126] As described above, in the voice processing method executed by the computer such as the processor 12, the first filter corresponding to the feature of the consonant is applied to each sub-string, that is, the voice signal is corrected by amplifying and emphasizing the frequency domain corresponding to the feature of the consonant, and the corrected voice signal is sent (output) to the speaker 2. Therefore, the user who hears the human voice output by the speaker 2 based on the corrected voice signal can easily hear the first phoneme of the word for each sub-string, and the intelligibility of the human voice is easily improved. Therefore, in such a voice processing method, there is an advantage that the user can easily hear the human voice regardless of the performance of the speaker 24 provided in the speaker 2.

[0127] In addition, in the embodiment, in the sound processing method, the sound signal is further corrected by applying a second filter corresponding to the type of the speaker 2, that is, by amplifying and emphasizing the power of the frequency domain with relatively small power based on the frequency characteristics of the human voice output by the speaker 2. Therefore, the human voice output by the speaker 2 is corrected according to the characteristics of the speaker 2, so there is an advantage that the user can more easily hear the human voice output by the speaker 2.

[0128] In addition, in the embodiment, in the sound processing method, by applying a third filter corresponding to the sound collection information, that is, amplifying and emphasizing the power in the frequency domain with a relatively large power based on the frequency characteristics of the sound around the speech device 2, the sound signal is further corrected. Therefore, the human voice output by the speech device 2 is corrected so as not to be masked by the sound around the speech device 2, and thus there is an advantage that the user can more easily hear the human voice output by the speech device 2.

[0129] [5. Other Embodiments]

[0130] As described above, the embodiments have been described, but the present disclosure is not limited to the above embodiments.

[0131] In the above embodiment, the processor 12 divides the generated string into one or more sub-strings based on syllables, but it is not limited thereto. For example, the processor 12 may also divide the generated string into one or more sub-strings based on words. As an example, when the generated string contains the string "せんたくき(sentakuki)", the processor 12 may also divide "せんたくき(sentakuki)" into one sub-string.

[0132] In addition, for example, the processor 12 may also divide the generated string into one or more sub-strings based on beats. As an example, when the generated string contains the string "せんたくき(sentakuki)", the processor 12 may also divide "せんたくき(sentakuki)" into five sub-strings: "せ(se)", "ん(n)", "た(ta)", "く(ku)", and "き(ki)".

[0133] In addition, for example, the processor 12 may also divide the generated string into one or more sub-strings based on the Chinese characters constituting the word. As an example, when the generated string contains the string "せんたくき(洗濯機)" (sentakuki (washing machine)), the processor 12 may also divide "せんたくき(sentakuki)" into three sub-strings: "せん(sen)", "たく(taku)", and "き(ki)".

[0134] In the above embodiment, when the sub-string contains the characteristics of multiple consonants, the filter corresponding to the characteristics of all the consonants is determined as the first filter, but it is not limited thereto. For example, the processor 12 may also determine the priority for the characteristics of multiple consonants for each sub-string in the sound signal.

[0135] As an example, the processor 12 may also determine as the first filter only the filter corresponding to the feature of the consonant with a high priority among the features of the multiple consonants. Here, high priority means that as long as the user can hear the consonant clearly, the user can understand the meaning of the string. In other words, the possibility that the user misidentifies the meaning of the string is low. In addition, the processor 12 may also determine as the first filter only the filter corresponding to the feature of the consonant with the highest priority among the features of the multiple consonants, or may determine as the first filter the filter corresponding to the features of the consonants whose priority order is up to a specified order.

[0136] In the above embodiment, the processor 12 performs the first filtering process, the second filtering process, and the third filtering process in the filtering process for correcting the sound signal, but the present invention is not limited thereto. For example, the processor 12 may not perform either the second filtering process or the third filtering process, or may not perform both of the processes.

[0137] In the above embodiment, the voice processing system 10 causes the speech device 2 to output a human voice in Japanese, but the present invention is not limited thereto. For example, the voice processing system 10 may cause the speech device 2 to output a human voice in another language such as English or Chinese. In this case, the processor 12 only needs to determine the first filter corresponding to the language of the human voice output by the speech device 2 for each sub-character string.

[0138] For example, in the case where the speech device 2 is made to output the voice of a person in English, the processor 12 divides the determined character string into one or more sub-character strings in units of words. Here, if a word contains multiple syllables, the processor 12 may further divide the determined character string into one or more sub-character strings in units of syllables. In addition, in this case, the processor 12, as in the case of Japanese, only needs to determine the first filter corresponding to the characteristics of the consonant (voice, nasality, persistence, irregularity, muteness and concentration) for each sub-character string. For example, words having the characteristic of the consonant of voice are "veal" and "feel". In addition, words having the characteristic of the consonant of nasality are "moot" and "boot". In addition, words having the characteristic of the consonant of persistence are "sheet" and "cheat". In addition, words having the characteristic of the consonant of irregularity are "sing" and "thing". In addition, words having the characteristic of the consonant of muteness are "weed" and "reed". For example, words having the consonant characteristic of concentration include "key" and "tea".

[0139] In the above embodiment, the processor 12 determines to make a pre-set speaker 2 among the multiple speaker devices 2 output the human voice when there are multiple speaker devices 2, but the present invention is not limited to this. For example, when there is a detection device capable of obtaining the position of the user in an environment where multiple speaker devices 2 are placed, the processor 12 may also determine the speaker device 2 that outputs the human voice based on the position of the user.

[0140] For example, it is assumed that one or more human sensors are provided in an environment where a plurality of speech devices 2 are placed, and information indicating speech devices 2 existing around the human sensors is stored for each human sensor in the memory of the processor 12. In this case, the processor 12 may obtain detection results from one or more human sensors via the communication I / F 11 (input I / F 11A), and determine to cause the speech device 2 corresponding to the human sensor indicating the presence of the user to output the human voice.

[0141] In the above embodiment, the communication I / F 11 of the server 1 serves as both the input I / F 11A and the output I / F 11B, but the present invention is not limited to this. For example, the input I / F 11A and the output I / F 11B may be different interfaces.

[0142] In addition, in the above-mentioned embodiment, the sound processing system 10 is implemented by a single device, but it can also be implemented as a plurality of devices. In the case where the sound processing system 10 is implemented by a plurality of devices, the functional components of the sound processing system 10 can be distributed to the plurality of devices in any manner. For example, the sound processing system 10 can also be implemented by being distributed in a plurality of servers. In addition, for example, the sound processing system 10 can also be implemented by being distributed in a server and a speech device. In addition, for example, the sound processing system 10 can also be implemented by only a speech device.

[0143] In addition, the communication method between the devices in the above-mentioned embodiment is not particularly limited. When two devices communicate with each other in the above-mentioned embodiment, a relay device (not shown) may exist between the two devices.

[0144] In addition, the order of the processing described in the above embodiment is an example. The order of multiple processing can also be changed, and multiple processing can also be performed in parallel. In addition, the processing performed by a specific processing unit can also be performed by other processing units. In addition, part of the digital signal processing described in the above embodiment can also be implemented by analog signal processing.

[0145] In addition, in the above-mentioned embodiments, each component can also be implemented by executing a software program suitable for each component. Each component can also be implemented by a program execution unit such as a CPU or a processor reading and executing a software program recorded in a recording medium such as a hard disk or a semiconductor memory.

[0146] In addition, each component may also be implemented by hardware. For example, each component may also be a circuit (or integrated circuit). These circuits may constitute one circuit as a whole, or they may be different circuits. In addition, these circuits may be general circuits or dedicated circuits.

[0147] In addition, the global or specific technical solutions of the present disclosure may also be implemented by a system, device, method, integrated circuit, computer program or a computer-readable recording medium such as a CD-ROM. In addition, it may also be implemented by any combination of a system, device, method, integrated circuit, computer program and recording medium. For example, the present disclosure may also be implemented as a sound processing method executed by a computer, or as a program for causing a computer to execute such a sound processing method. In addition, the present disclosure may also be implemented as a computer-readable non-temporary recording medium having such a program recorded thereon. In addition, the program here includes an application program for enabling a general-purpose information terminal to function as the sound processing system of the above-mentioned embodiment.

[0148] In addition, various modifications that can be conceived by those skilled in the art to the embodiments or embodiments achieved by arbitrarily combining the components and functions of the embodiments without departing from the gist of the present disclosure are also included in the present disclosure.

[0149] (Summarize)

[0150] As described above, in the sound processing method related to the first technical solution, event information related to the event is obtained from the information source device 3 or the information source service 4 (S1); based on the obtained event information, a character string that causes the speaking device 2 to speak is determined (S4); the determined character string is divided into one or more sub-character strings (S5); a sound signal is generated based on the character string (S7); for the generated sound signal, filtering processing is performed by applying the first filter corresponding to the characteristics of the consonant to each sub-character string, and the sound signal is modified (S6, S8); and the modified sound signal is output (S11).

[0151] This has the advantage that the user can easily hear the human voice regardless of the performance of the speaker 24 included in the speaker device 2.

[0152] Furthermore, in the speech processing method according to the second technical means, in the first technical means, the character string is divided into one or more sub-character strings based on syllables.

[0153] This has the advantage of ensuring the intelligibility of human voice and reducing the load of processing for correcting the audio signal compared to the case of dividing the character string character by character.

[0154] Furthermore, in the speech processing method according to the third aspect, in the first or second aspect, in the filtering process, a second filter corresponding to the type of the speaker 2 is further applied to the speech signal (S9).

[0155] Thus, the human voice outputted by the speaker 2 is corrected according to the characteristics of the speaker 2 , so there is an advantage that the user can more easily hear the human voice outputted by the speaker 2 .

[0156] In addition, in the sound processing method related to the fourth technical scheme, in any one of the first to third technical schemes, sound collection information (S3) obtained by collecting sounds around the speaking device 2 is obtained; in the filtering process, the third filter (S10) corresponding to the above-mentioned sound collection information is also applied to the above-mentioned sound signal.

[0157] Thus, the human voice outputted from the speech device 2 is corrected so as not to be masked by the surrounding sounds of the speech device 2 , so there is an advantage that the user can more easily hear the human voice outputted from the speech device 2 .

[0158] Furthermore, the program according to the fifth aspect causes one or more processors to execute the sound processing method according to any one of the first to fourth aspects.

[0159] This has the advantage that the user can easily hear the human voice regardless of the performance of the speaker 24 included in the speaker device 2.

[0160] In addition, the sound processing system 10 related to the sixth technical solution includes: a processor 12 that modifies the sound signal; and an output I / F 11B that outputs the modified sound signal. The processor 12 is an example of a signal processing circuit. The processor 12 determines a string that causes the speech device 2 to speak based on the acquired event information, divides the determined string into one or more sub-strings, generates a sound signal based on the string, and performs filtering processing on the generated sound signal by applying a first filter corresponding to the characteristics of the consonant to each sub-string, thereby modifying the sound signal.

[0161] This has the advantage that the user can easily hear the human voice regardless of the performance of the speaker 24 included in the speaker device 2.

[0162] Industrial Applicability

[0163] The sound processing method disclosed herein can be applied to a system or the like that processes human voice for reproduction by a speaker.

[0164] Description of symbols

[0165] 1 Server

[0166] 10 Sound Processing System

[0167] 11Communication I / F

[0168] 11A Input I / F

[0169] 11B output I / F

[0170] 12 Processor

[0171] 13 Memory

[0172] 14 Storage Devices

[0173] 2 Speaking equipment

[0174] 21 Communication I / F

[0175] 22 Processor

[0176] 23 Memory

[0177] 24 Speakers

[0178] 25 Microphone

[0179] 3 Information source equipment

[0180] 4 Information source service

[0181] N1 Network

Claims

1. A sound processing method, wherein: Obtain event information related to the event from the information source device or information source service, Based on the event information obtained, a character string is determined to cause the speech device to speak. The determined character string is divided into one or more sub-character strings. Generate a sound signal based on the above string, The generated audio signal is subjected to filtering processing by applying a first filter corresponding to the characteristics of the consonants to each of the sub-character strings, thereby modifying the audio signal. The corrected sound signal is output.

2. The sound processing method according to claim 1, wherein: The character string is divided into the one or more sub-character strings based on syllables.

3. The sound processing method according to claim 1 or 2, wherein: In the filtering process, a second filter corresponding to the type of the speaker is further applied to the audio signal.

4. The sound processing method according to claim 1 or 2, wherein: Acquiring sound information obtained by collecting sounds around the above-mentioned speaking device, In the filtering process, a third filter according to the collected sound information is further applied to the audio signal.

5. A program, wherein: One or more processors are used to execute the sound processing method according to claim 1 or 2.

6. A sound processing system, wherein: have: An input interface for obtaining event information related to an event from an information source device or an information source service; A signal processing circuit to modify the sound signal; and Output interface, outputs the corrected sound signal. The above signal processing circuit, Based on the event information obtained, a character string is determined to cause the speech device to speak. The determined character string is divided into one or more sub-character strings. Generate a sound signal based on the above string, The generated audio signal is corrected by performing a filtering process of applying a first filter corresponding to the characteristics of the consonant for each of the sub-character strings to the generated audio signal.

Images