JP4134844B2 - Hearing aids - Google Patents

Description

  The present invention relates to a hearing aid apparatus that prevents excessive hearing of environmental sound in a non-speech section and improves the ease of hearing of a sound in a speech section.

  Hearing aids are devices that assist the hearing of people who have difficulty hearing. Conventional hearing aids, regardless of the type of collected external sound, that is, whether the external sound is a sound (a sound emitted from the mouth to convey the intention of a human being) or a sound other than sound, is uniform. Were amplified with the same amplification factor and output. However, users of hearing aids sometimes feel uncomfortable because they can hear not only the other party's voice (sound) but also surrounding sounds (environmental sounds). For example, the sounds of tableware hitting each other when washing dishes and the sound of dropping dishes are very harsh to the hearing aid user.

Therefore, conventionally, a hearing aid has been proposed in which the amplification factor is increased only when a voice is detected, and the amplification factor is not increased when no voice is detected (see, for example, Patent Document 1). This hearing aid reduces the gain of the amplifier when the relative noise level is high, based on the ratio between the power level of the input sound and the power level of the sound in the silent period. Thereby, it is possible to prevent an unpleasant sound due to excessively amplified noise from being added to the user of the hearing aid.
JP-A-8-294197 (pages 4-11, 1-7)

  The hearing aid disclosed in Patent Document 1 acquires the number of short-time zero crossings to detect whether or not it is a silent section, and accumulates zero-crossings at set time intervals to acquire the number of short-time zero crossings. The process of doing. Therefore, in this hearing aid, even if the set time is set to a short time, until a sound is output by executing a series of processes such as acquisition of the number of short-time zero crossings of the input sound, determination of the voice interval, and amplification of the voice, A time of at least about 30 to 100 msec was always necessary.

  However, humans have the property of feeling uncomfortable when there is a difference of 15 msec or more between the movement of the speaker's mouth and the voice of the speaker (the content of the speech). In addition, a user of a hearing aid usually listens to a speaker's speech while watching the movement of the speaker's mouth in order to grasp the content of the speaker's speech. Therefore, there is a problem that the user of the hearing aid described in Patent Document 1 must listen to the speaker's story while always feeling uncomfortable that the movement of the speaker's mouth and the voice of the speaker are out of sync. there were.

  Therefore, in order to solve the above problem, the present invention prevents an environmental sound from being excessively heard in a non-voice section, and uses a sense of incongruity due to a difference between the movement of the speaker's mouth and the voice of the speaker. It is an object of the present invention to provide a hearing aid device in which the voice is appropriately amplified in the voice section and the ease of hearing is improved with little person holding.

  The present invention has the following configuration as means for solving the above problems.

(1) sound collecting means for collecting foreign sounds;
Buffer means for temporarily storing foreign sound data collected by the sound collecting means;
Writing means for writing to the buffer means while sequentially updating the data of the external sound collected by the sound collecting means;
Determining means for determining whether or not there is sound in the external sound data for a predetermined period stored in the buffer means;
Means for reading external sound data from the buffer means, and when the normal determination means determines that there is sound, reads the external sound data immediately after writing by the writing means;
When the determination result of the determination means changes from no sound to presence of sound , the reading position of the external sound data read from the buffer means is determined as the external position determined by the determination means after a certain period of time has elapsed from the writing by the writing means. Reading means for jumping to the position of the sound data and gradually shifting from the determined position of the external sound data to the position of the external sound data immediately after the writing;
Amplifying means for amplifying and outputting the external sound data determined by the determining means with the sound and the external sound data determined by the determining means with no sound at different amplification rates;
When the determination by the determination unit changes from no sound to sound, notification means for notifying that effect;
It is provided with.

In this configuration, the hearing aid apparatus reads out the determined external sound data with sound from the buffer means when detecting the sound that is a sound emitted from the mouth in order to convey the intention of the person. If the person determines that there is sound, the amplification factor is set higher than when the sound is determined not to exist, so that when the sound is detected, the emphasized sound can be output from the beginning of the sound. . In addition, when the sound is detected, the determined external sound data with the sound is read from the buffer means and gradually shifted to the external sound data immediately after the writing, so that the user of the hearing aid device outputs the emphasized sound. You can hear without leaking from the beginning. Further, the external sound is not output in real time only during a predetermined period after the voice is detected, and the external sound can be heard almost in real time in most other periods. In addition, the user of the hearing assistance device can grasp that the sound is output from the hearing assistance device by the notification means, so that the voice of the speaker can be concentrated and heard from the beginning. Note that, as the notification means, a speaker that notifies the user of the hearing assistance device by sound, an LED that notifies by light, a vibration motor that notifies by vibration, and the like are suitable.

  (2) When the reading means shifts the external sound data read from the buffer means, the reading means includes speech speed conversion means for converting the speech speed of the external sound data in accordance with the shift.

  In this configuration, when sound is detected, it is possible to read out the determined external sound data having the sound from the buffer means and gradually shift the speech speed to the external sound data just after writing. Therefore, even if the sound gradually shifts when the hearing aid apparatus detects the sound, the user can hear this sound without a sense of incongruity.

  (3) When the determination result of the determination unit changes from the presence of sound to the absence of sound, the amplifying unit reduces the amplification factor at a predetermined rate to the amplification factor for the external sound determined to have no sound. Is amplified.

  In this configuration, if the external sound collected by the hearing aid device after the speaker finishes speaking is not included in the sound and only the environmental sound is present, the amplification factor of the external sound output from the hearing aid device is set to no sound. It can be reduced at a predetermined rate to the amplification factor for the external sound determined to be. Therefore, when the hearing aid device does not detect the sound for a while after detecting the sound, the volume of the environmental sound is gradually lowered at a predetermined rate, so that the user of the hearing aid device ends the voice of the speaker. Even if the hearing assistance device no longer detects the sound, the foreign sound can be heard without feeling uncomfortable with the change in the volume of the external sound output from the hearing assistance device. In addition, since the hearing aid apparatus reduces the amplification rate of the external sound by a predetermined rate, if the amplification rate is decreasing even if the speaker resumes speaking after a short break, the external sound will not be heard. Since it is not suddenly greatly amplified, the user of the hearing aid device does not have to be surprised if a loud sound is suddenly heard.

  The hearing aid apparatus according to the present invention does not allow excessive environmental sounds to be heard in non-speech intervals, and there is almost no sense of incongruity due to the difference between the mouth movement of the speaker and the speech of the speaker. There is an advantage that the sound can be easily heard after being amplified.

  Hereinafter, a hearing aid will be described as an example of a hearing aid apparatus. FIG. 1 is a block diagram showing a schematic configuration of a hearing aid. Here, in the following description, the voice refers to a sound emitted from the mouth in order to convey a will, and the environmental sound refers to a sound that does not include the voice.

  The hearing aid 1 includes a microphone 2, a signal processing unit 3, and a speaker 4. The signal processing unit 3 includes an A / D converter 11, a DSP 12, a CPU 13, an operation unit 14, a D / A converter 15, a memory 16, and an input / output interface 17.

  The microphone 2 collects the external sound around the user of the hearing aid 1 and outputs the collected analog sound signal to the signal processing unit 3. The signal processing unit 3 performs signal processing such as speech speed conversion and amplification, and outputs the result to the speaker 4. The speaker 4 converts the audio signal signal-processed by the signal processing unit 3 into air vibration and emits it to the user's ear canal.

  In the signal processing unit 3, the A / D converter 11 converts the analog sound signal output from the microphone 2 into a digital sound signal. The DSP 12 performs sound processing such as speech speed conversion processing and amplification processing on the digital sound signal output from the A / D converter 11. The CPU 13 controls each part of the hearing aid 1. The operation unit 14 is for performing various settings such as volume setting, time setting, and amplification factor setting of the hearing aid 1. The D / A converter 15 converts the digital sound signal processed by the DSP 12 into an analog sound signal. The memory 16 stores a setting value of the hearing aid 1 set by the user. The input / output interface 17 is an interface for exchanging signals with the personal computer 8.

  Next, the configuration of the DSP 12 will be described. FIG. 2 is a block diagram showing a schematic configuration of a DSP built in the hearing aid. The DSP 12 includes a write controller 21, a ring buffer 22, a read controller 23, a voice section detection unit 24, a read controller 25, a speech speed conversion unit 26, a gain controller 27, and a notification sound generation unit 28.

  The write controller 21 sequentially records the digital sound signal sent from the A / D converter 11 in the ring buffer 22 in accordance with the write address indicated by the pointer that advances in the sampling period.

  The ring buffer 22 sequentially stores the digital sound data output from the write controller 21, and when the memory area is filled with stored data, the old data is overwritten in order from the oldest data to delete the old data. In place of the ring buffer, a shift register or a FIFO memory may be used.

  The read controller 23 sequentially reads out 50 msec of the digital sound data stored in the ring buffer 22 in the non-speech section in which only the environmental sound is detected, and outputs the digital sound data to the speech section detection unit 24. Here, the digital sound data for 50 msec read from the ring buffer 22 by the read controller 23 or the time of 50 msec is also referred to as a frame hereinafter.

  The voice section detection unit 24 analyzes the digital sound data read from the ring buffer 22 and determines whether or not voice is included in the digital sound data. This analysis takes 50 msec. When it is determined that the voice is included, the voice section detection unit 24 outputs a voice detection signal to the speech speed conversion unit 26, the gain controller 27, and the notification sound generation unit 28. If it is determined that no voice is included, an undetected signal is output to the speech speed conversion unit 26, the gain controller 27, and the notification sound generation unit 28. Note that the voice section detection unit 24 analyzes whether or not a voice is included by using well-known techniques such as VAD (Voice Activity Detection), LPC analysis, and cepstrum method.

  The read controller 25 reads the digital sound data from the ring buffer 22 and outputs it to the speech speed conversion unit 26 in order to output the environmental sound and sound from the speaker 4. That is, the read controller 25 reads the digital sound data in accordance with the read address indicated by the pointer that advances at the sampling period in the non-voice section, and outputs the digital sound data to the speech speed conversion unit 26 in the non-voice section. Here, in the non-voice section, the digital sound data read by the read controller 25 has almost no delay with respect to the digital sound data written by the write controller 21.

  When the speech section detection unit 24 detects the start of the speech section, the speech speed conversion unit 26 outputs a signal instructing the read controller 25 to jump the read address backward. In addition, the speech speed conversion unit 26 instructs the read controller 25 to increase the stepping speed of the read address for a predetermined time after the voice segment detection unit 24 detects the start of the voice segment. When the speech speed conversion unit 26 issues an instruction to increase the stepping speed, the voice pitch output from the speaker 4 increases and the voice speed increases. If the supplementary conversion unit 26 is provided with a known variable speed reproduction function, it is possible to change only the speed without changing the pitch (key) when changing the stepping speed. In addition, when the speech speed conversion unit 26 issues an instruction to increase the stepping speed, the speech (foreign sound) data output from the read controller 25 is a well-known silent section or long vowel section. It is also possible to use a method in which the time compression is performed to convert the speech speed of the voice data. When the time compression is set, the sound (output sound) output from the hearing aid 1 is a foreign sound collected by the microphone 2 with the sound pitch and sound speed, although the stepping speed is increased. Almost the same as (input sound).

  The gain controller 27 changes the amplification factor according to the signal sent from the voice section detector 24 and amplifies the digital sound data sent from the speech speed converter 26. That is, when an undetected signal is sent from the voice section detector 24, the gain controller 27 amplifies the digital sound data sent from the speech speed converter 26 with an amplification factor set for environmental sounds. Further, when a voice detection signal is sent from the voice section detection unit 24, the gain controller 27 converts the digital sound data (voice data) sent from the speech speed conversion unit 26 with an amplification factor set for voice. Amplify to output enhanced sound. Here, the amplification factor of the environmental sound and the amplification factor of the sound can be set according to the use environment of the user. For example, the gain of the environmental sound of the gain controller 27 is set to a value that amplifies the external sound (environmental sound) to such an extent that the user can notice a loud sound, and the sound gain is determined by the user. It should be set to a value that amplifies the external sound (sound) to such an extent that it can be easily heard.

  When the sound detection signal is sent from the sound section detection unit 24, the notification sound generation unit 28 generates a sound for notifying the user of the hearing aid 1 that the sound is output and outputs the sound to the D / A converter 15. To do. The tone color and playback time of the notification sound can be set in advance according to the user's preference.

  Next, the overall operation of the hearing aid 1 will be described. FIG. 3 is a timing chart for explaining the operation of the hearing aid. FIG. 4 is a diagram for explaining the sound detection operation of the external sound data in the hearing aid. FIG. 5 is a diagram showing the relationship between the input pointer and the output pointer of the ring buffer at each timing shown in FIG. During the operation of the hearing aid 1, the input pointer and the output pointer of the ring buffer 22 move sequentially, but the position of the input pointer is fixed in FIG. 5 so that the operation of the ring buffer 22 can be easily understood. Show.

  The hearing aid 1 constantly writes the external sound collected by the microphone 2 in the ring buffer, and reads the external sound data for a certain period (50 msec) from the external sound data stored in the ring buffer 22 to determine the presence or absence of sound. To do. Then, according to the determination result, the external sound data is read from the speaker 4 by changing the reading position and amplification factor of the external sound data.

  (1) The hearing aid 1 can switch between the amplification factor for environmental sound and the amplification factor for sound to amplify and output the external sound. In the initial state, the sound is amplified with the amplification factor for the environmental sound. Is set to When the power is turned on, the hearing aid 1 collects the external sound with the microphone 2 and writes the external sound data in the ring buffer 22 and reads the external sound data immediately after writing to amplify the environmental sound. Amplify at a rate and output external sound from the speaker 4. Therefore, in the ring buffer 22, the output pointer (for the read controller 25) and the input pointer (for the write controller 21) move at the same speed as shown in FIG. Output from the speaker 4 in real time.

  (2) When the external sound data for a certain period (50 msec) is accumulated in the ring buffer 22, the hearing aid 1 analyzes and determines whether or not the sound is included in the external sound data for a certain period of time. After that, every time foreign sound data is accumulated for a certain period, the analysis / determination of the presence / absence of sound is continued.

  That is, as shown in FIG. 4, the hearing aid 1 sequentially writes the input sound data (foreign sound data) as digital sound data in the ring buffer 22 in the non-speech period, and reads out the external sound data immediately after the writing, The external sound is output from the speaker 4 after being amplified with the amplification factor for environmental sound. When the external sound data (first frame) for one frame (50 msec) is accumulated in the ring buffer 22, the read controller 23 reads out the external sound data of the first frame and outputs it to the audio section detection unit 24. The voice section detection unit 24 analyzes for about 50 msec whether or not voice is included in the external sound data of the first frame. Then, the voice section detection unit 24 outputs an undetected signal to the speech speed conversion unit 26 and the gain controller 27 when it is determined that the sound is not included in the external sound data after the analysis is completed. As a result, the speech speed conversion unit 26 outputs the external sound data to the gain controller 27 without performing the speech speed conversion, and the gain controller 27 amplifies the external sound with the amplification factor for the environmental sound without changing the amplification factor. To do. Therefore, the hearing aid 1 outputs environmental sound (foreign sound) in real time without time delay.

  As shown in FIG. 3, when the speaker speaks from time t1 to time t3, the hearing aid 1 performs the following operation.

  (3) When the previous determination result is no sound but the current determination result is sound, the hearing aid 1 emphasizes from the beginning of the sound and outputs it. The sound data is read from the ring buffer 22, amplified with a sound amplification factor, and output from the speaker 4.

  In the hearing aid 1, for example, when the external sound data for 50 msec from the time t 1 when the speaker starts speaking is accumulated in the ring buffer 22, the read controller 23 reads the external sound data and sends it to the voice section detection unit 24. Data is output, and the voice section detector 24 analyzes the presence or absence of voice for about 50 msec to determine the presence or absence of voice. When the voice section detection unit 24 determines that there is voice, the voice section detection unit 24 outputs a voice detection signal to the speech speed conversion unit 26 and the gain controller 27. When detecting the voice detection signal, the speech speed conversion unit 26 outputs a pointer position change signal to the read controller 25. Further, when the gain controller 27 detects the sound detection signal, the gain controller 27 changes the amplification factor to amplify the external sound with the sound amplification factor. Therefore, in the ring buffer 22, the output pointer jumps 100 msec before, and the input pointer is positioned at an address advanced by 100 msec of the digital sound data from the output pointer as shown in FIG. Therefore, immediately after the hearing aid 1 detects the sound (time t2), the sound spoken by the speaker about 100 msec before is amplified by the amplification factor for the sound and output from the speaker 4.

  (4) As described above, the user of the hearing aid 1 feels uncomfortable when the time lag between the movement of the mouth of the speaker and the sound output from the hearing aid 1 is 15 msec or more. Therefore, in order to solve this problem, the hearing aid 1 continues to read the audio data from the ring buffer 22 so that the sound is not interrupted while continuing the time compression of the sound using the speech speed conversion technique. For this reason, in the ring buffer 22, the audio data is quickly read, and the output pointer gradually approaches the input pointer as shown in FIG.

  (5) When the speaker continues speaking, when 1 sec to 5 sec elapses from the start of the speech speed conversion, the output pointer catches up with the input pointer in the ring buffer 22 and the hearing aid 1 outputs as shown in FIG. There is no time delay in the voice (time t5). The hearing aid 1 stops time compression when there is no time delay in the sound to be output, reads out the external sound data immediately after being written to the ring buffer 22, amplifies it with the amplification factor for sound, and outputs it from the speaker 4. In the ring buffer 22, the output pointer and the input pointer move at the same speed until (8).

  (6) When the speaker continues speaking, the voice section detection unit 24 outputs a voice detection signal to the gain controller 27 and amplifies the external sound (voice) data to the gain controller 27 with the amplification factor for voice. Let

  (7) When the speaker stops speaking and no sound is included in the foreign sound, the voice section detection unit 24 determines that there is no voice 100 msec (time t4) after the speaker stops speaking (time t3), A detection signal is output to the gain controller 27.

  (8) In the hearing aid 1, when the state in which no sound is detected continues for a while, the gain controller 27 gradually reduces the amplification factor of the external sound from the sound value to the environmental sound value. For example, when a state in which no sound is included in the external sound continues for 1 minute, the gain controller 27 reduces the amplification rate of the external sound by 20%. Further, if the state in which the sound is not included in the external sound continues, the gain controller 27 decreases the amplification factor by 20% every 1 minute until the amplification factor of the external sound becomes the amplification factor for the environmental sound. Let The amplification factor may be decreased stepwise as described above, or may be decreased linearly at a predetermined rate, and the decrease rate can be set to an arbitrary value. When the hearing aid 1 detects that the sound is included in the external sound data when the speaker resumes speaking while the amplification rate of the external sound is decreasing, the process (3), that is, the sound The time compression (speech speed conversion) and the change of the amplification factor to the voice amplification factor are immediately started to emphasize the voice.

  Further, in the hearing aid 1, when the reduction of the amplification rate of the external sound is completed, the processing described in (2) is performed.

  In the hearing aid 1, when the start of the speech section is detected, the ring buffer 22 is read back to just before the start point, so that when the speaker starts speaking, the voice of the speaker is not leaked from the beginning of the speech. Can hear. In addition, the hearing aid 1 has a time delay of 100 msec between the input sound and the output sound immediately after detecting the sound as described above, but the time delay is eliminated after a few seconds due to the speech speed conversion. Can hear foreign sounds using the hearing aid 1 with almost no sense of incongruity.

  In addition, when the voice detection is completed, the voice amplification rate is gradually reduced. Therefore, when the speaker keeps speaking after taking a breath, the voice amplification rate does not fluctuate greatly. You can listen to the sound without any discomfort.

  Here, as described in the above (3), when the speech section detection unit 24 finishes the analysis and determines that the sound is included in the external sound data, the speech section detection unit 24 generates a notification sound. It is also possible to output the sound detection signal to the unit 28 so that the notification sound generation unit 28 generates the notification sound and then outputs the sound. When the sound detection signal is detected, the notification sound generator 28 generates a notification sound signal and outputs this signal to the D / A converter 15 for about 50 msec, for example, so as to output the notification sound from the speaker 4. In this case, the speech speed conversion unit 26 may read the external sound data including the sound from the ring buffer 22 after the notification sound generation unit 28 outputs the notification sound signal.

  In the above description, the notification sound generation unit has been described as an example of notification means for notifying that sound is output. However, the present invention is not limited to this, and may be in other forms. good. For example, in the case of a glasses-type hearing aid or the like, an LED may be provided at a position where the user of the hearing aid can easily check, and this LED may blink to notify that a sound is output. In the case of an ear-hook type or pocket-type hearing aid, a vibration motor may be provided inside to notify that sound is output by vibration.

  In the hearing aid 1, the time for analyzing / detecting the presence / absence of speech, the time for reducing the amplification factor of speech, the time for performing speech speed conversion, and the like can be changed according to the preference of the user of the hearing aid 1. it can. Further, the environmental sound amplification factor and the sound amplification factor of the hearing aid 1 can also be changed according to the hearing ability and preference of the user.

  Further, in the hearing aid 1, the user of the hearing aid 1 uses the personal computer 8 because the input / output interface 17 of the hearing aid 1 is connected to the personal computer 8 in order to adjust various settings. Thus, it is possible to easily adjust and change various settings such as setting of the environmental sound and the gain of the sound of the gain controller 27.

  In the above description, a hearing aid has been described as an example of an embodiment of a hearing aid, but the present invention is not limited to this. For example, a device that outputs an image and sound of a speaker at the same time, such as a video phone, is built in with the hearing aid device of the present invention, and the sound is output depending on whether the sound output from this device is sound or not. It is preferable to adopt a configuration in which the gain of the voice is changed or the speech speed is converted. Thereby, the voice of the speaker can be emphasized, and the hearing impaired person can easily use this apparatus.

It is the block diagram which showed schematic structure of the hearing aid. It is the block diagram which showed schematic structure of DSP which a hearing aid contains. It is a timing chart for demonstrating operation | movement of a hearing aid. It is a figure for demonstrating the audio | voice detection operation | movement of the external sound data in a hearing aid. It is the figure which showed the relationship between the input pointer and output pointer of a ring buffer in each timing shown in FIG.

Explanation of symbols

1- hearing aid 1, 2-microphone, 3-signal processing unit, 4-speaker,
11-A / D converter, 12-DSP, 13-CPU, 14-operation unit,
15-D / A converter, 16-memory, 17-input / output interface

Claims (3)

Sound collecting means for collecting foreign sounds;
Buffer means for temporarily storing foreign sound data collected by the sound collecting means;
Writing means for writing to the buffer means while sequentially updating the data of the external sound collected by the sound collecting means;
Determining means for determining whether or not there is sound in the external sound data for a predetermined period stored in the buffer means;
Means for reading external sound data from the buffer means, and when the normal determination means determines that there is sound, reads the external sound data immediately after writing by the writing means;
When the determination result of the determination means changes from no sound to presence of sound , the reading position of the external sound data read from the buffer means is determined as the external position determined by the determination means after a certain period of time has elapsed from the writing by the writing means. Reading means for jumping to the position of the sound data and gradually shifting from the determined position of the external sound data to the position of the external sound data immediately after the writing;
Amplifying means for amplifying and outputting the external sound data determined by the determining means with the sound and the external sound data determined by the determining means with no sound at different amplification rates;
When the determination by the determination unit changes from no sound to sound, notification means for notifying that effect;
A hearing aid device comprising:   The hearing aid apparatus according to claim 1, further comprising: speech speed conversion means for converting the speech speed of the external sound data in accordance with the shift when the reading means shifts the external sound data read from the buffer means.   The amplification means amplifies the external sound data while decreasing the amplification factor at a predetermined rate to the amplification factor for the external sound determined to have no sound when the determination result of the determination means changes from the presence of the sound to the absence of the sound. The hearing aid apparatus according to claim 1 or 2.

Images