JP3351745B2 - Hearing aid with pitch adjustment function - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a hearing aid having a pitch conversion function for compensating for hearing loss that has deteriorated due to aging or compensating for hearing loss of a hearing-impaired person.

[0002]

2. Description of the Related Art Conventionally, this type of hearing aid is a hearing aid for a hearing-impaired person having a high-pitched sound-fall type hearing characteristic. (Also called frequency compression). If the hearing person hears the sound that has been subjected to such processing, it only sounds very low.
However, the hearing impaired person can hear the unheard voice (although the voice is lowered), so that it is possible to understand the contents of the language.

FIG. 9 shows an example of a hearing aid using a conventional PARCOR analysis / synthesis system. In FIG. 9, reference numeral 10 denotes an acoustic signal input terminal for AD-converting the microphone output of the hearing aid, and 1
Is an AD converter that converts an analog signal into a digital signal at a sampling frequency f _si Hz, 2 is a PARCOR coefficient (K _i ), a voiced / unvoiced determination result (U / V), a pitch frequency (P _i ), and a sound source amplitude coefficient ( PARCOR that outputs each parameter of A)
The analyzing unit 3 is a PARCOR synthesizing unit that re-synthesizes the audio signal using the parameters output from the PARCOR analyzing unit 2, and the PARCOR synthesizing unit 3 controls the ratio of converting the frequency spectrum.
The control unit 5 outputs a conversion coefficient (C) to the digital signal D at a sampling frequency f _so Hz.
The A conversion unit 20 is an output terminal for extracting a sound whose frequency spectrum has been converted, and is configured to amplify an analog signal output from the output terminal 20 so that the analog signal can be heard with an earphone or the like.

[0004] The hearing aid configured as described above operates as follows. The PARCOR analysis unit 2 extracts each parameter by multiplying the audio signal by a window function of about 30 ms, and moves the window function by a certain time length (T _i ms: generally called a frame shift or a frame period). Then, the parameter extraction process is continued. PARCOR synthesizing unit 3, based on the conversion coefficient supplied from the control unit 4 (C) calculates the frame period at the time of re-synthesis _{(T o ms), T o} ms using the parameters obtained by the PARCOR analysis section 2 The audio signal is re-synthesized every time. The DA converter 5 converts the resynthesized audio signal into an analog signal at a sampling frequency f _so Hz different from that of the AD converter 1.

The conventional example operates as described above.
And the sampling frequency f of the AD converter 1 _si Hz and analysis frequency
Lame period T_i ms is determined at the time of design and the conversion factor
(C), composite frame period T_o ms and DA converter 5
Sampling frequency f_so Hz is calculated using equation (1) during operation.
Determined. C = f_so/ F_si= T_o/ T_i ... (1) For example, 50% of the frequency spectrum of the voice band 5 kHz
(C = 0.5) When performing compression (frequency conversion), f_si= 1
0 kHz, T_i= 10 ms, then f_so= 0.5 × f_si= 5 kHz T_o= 0.5 × T_i= Compress the frequency spectrum using the value of 5 ms (frequency conversion)
You. Therefore, the audio signal immediately before DA conversion is
It is reduced to 0.5 times the time length of the signal. This is A
DA conversion is performed at a frequency 0.5 times the sampling frequency of the D conversion unit 1.
By switching, it returns to exactly the same time length as the input voice
be able to. Moreover, at 0.5 times the sampling frequency, DA
Because of the conversion, the audio band is compressed 0.5 times (frequency
Conversion).

[0006]

The above-mentioned conventional frequency spectrum compression type hearing aid uses a PARCOR type audio signal model. However, only the audio signal is not always input to the microphone of the hearing aid, and in a normal living environment, “sound” other than voice, such as living noise, is mixed and input from the microphone. Therefore,
The pitch extraction processing in PARCOR analysis unit 2 does not work properly,
On the contrary, there is a problem that the sound becomes difficult to hear. Further, there is a problem that it is impossible to express the noise component by the audio signal model, and there is a problem that an unusual sound is mixed in the re-synthesized speech, thereby giving a user discomfort.

[0007] The present invention solves the above-mentioned conventional problems, and enables the frequency spectrum conversion of all "sounds" input from a microphone of a hearing aid without limiting to an audio signal, thereby reducing discomfort. An object of the present invention is to provide an excellent hearing aid that is extremely easy to hear.

[0008]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problem, the present invention provides a method for performing AD processing on an acoustic signal input from a microphone of a hearing aid without using a signal model, and then performing interpolation processing to obtain a sample number. The frequency spectrum is converted by increasing the time length, and the time length is returned to the original length using the PICOLA method or the like. According to the present invention, it is possible to listen not only by simply converting the pitch frequency and frequency spectrum of the audio signal but also by lowering the frequency of all “sounds” such as musical sounds and environmental noises, that is, by lowering the pitch. Further, the present invention has a function of adjusting the degree of adjusting the pitch to the most audible state in accordance with the pitch of the speaker in addition to the function of converting the pitch. Therefore, according to the present invention, an excellent hearing aid effect that is extremely easy to hear with little discomfort can be obtained.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A hearing aid according to a first aspect of the present invention comprises a means for extending a time length by interpolating a discrete audio signal, and a periodic waveform synchronized with a period in a periodic audio signal section. Means for correcting the time length by thinning out a signal having the same cycle length as that of a periodic signal section in the non-periodic sound signal section, and a pitch for determining the time length elongation rate of the discrete sound signal. It has a conversion rate adjusting means so that pitch conversion can be performed without depending on an input signal, and since an abnormal sound is not mixed, an excellent hearing aid effect that is extremely easy to hear with little discomfort can be obtained. Has features.

Further, the hearing aid according to the second aspect of the present invention has means by which the user of the hearing aid can adjust the pitch by manual operation, so that the user can judge the state of the input sound signal and listen most easily. It has the characteristic that it can be adjusted to a pitch.

Further, the hearing aid according to the third aspect has a means for automatically adjusting the pitch so that the voice can be heard most easily. No adjustment is performed, and in the case where the input voice is the voice of a child or a high-pitched female, the voice is automatically converted to a low voice.

Further, the hearing aid according to the fourth aspect automatically adjusts the pitch so that the pitch of the voice is the most audible, and automatically converts the pitch so that the voice quality does not change unnaturally. Since the means for limiting the rate is provided, it is possible to prevent the pitch conversion rate from becoming too large, thereby preventing loss of the voice of the speaker.

Further, the hearing aid according to the fifth aspect has means for adjusting the pitch conversion rate by a user's manual operation prior to the means for automatically adjusting the pitch conversion rate.
The feature is that the user can determine the state of the input sound signal and give priority to the automatic adjustment function of the pitch conversion rate so that the user can manually adjust the pitch so that the pitch becomes the most audible.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. (Embodiment 1) FIG. 1 shows a configuration of a hearing aid according to a first embodiment of the present invention. In FIG. 1, reference numeral 101 denotes an audio signal input terminal to which an output signal of a microphone of a hearing aid is connected. Reference numeral 11 denotes AD conversion means for converting an audio signal coming from the microphone into a discrete audio signal at a sampling frequency fs Hz. Reference numeral 21 denotes a discrete audio signal interpolated to increase the number of samples (the number of data). Acoustic signal interpolation means for extending the time length of the audio signal, 31 is a time length correction means for performing waveform thinning-out processing to return the discrete audio signal to the original time length, and 41 is a discrete audio signal interpolation means 21 Pitch conversion rate adjusting means for outputting a conversion coefficient (C) for defining the time expansion rate of the dynamic audio signal;
Is a DA conversion means for converting the output signal of (1) into an analog signal at a sampling frequency fs Hz. Reference numeral 201 denotes a terminal that outputs a pitch conversion signal. The analog signal output from the output terminal 201 is amplified so that the analog signal can be received by an earphone or the like, thereby constituting a hearing aid.

The operation of the hearing aid with the pitch conversion adjusting function configured as described above will be described below. In FIG. 1, a pitch conversion rate adjusting means 41 converts a pitch conversion coefficient (C) into an acoustic signal interpolation means 21 and a time length correction means 3.
1 to output. For example, when a value of C = 0.5 is output to the audio signal interpolation means 21, the audio signal interpolation means 21 interpolates between the sample values to obtain fs / C Hz, that is, 2
Resample at twice the sampling frequency. Means for interpolating between sample values include linear interpolation, quadratic curve interpolation, spline interpolation,
Alternatively, a method of converting the data into a frequency axis and performing processing is used.

By the way, when resampling is performed, the number of sample values (data) increases by fs / C times (doubled in the above example). When the resampled data is converted to an analog signal at a sampling frequency of fs / C Hz (2 fs Hz in the above example), the pitch does not change, and the reproduction time length is the same as the time length of the input acoustic signal. It will be long. However, in the present invention,
Since the sampling frequency is converted to an analog signal by the DA conversion means 51 having the same fs Hz as that at the time of the AD conversion, the reproduction time length is increased by the increased data,
Therefore, an acoustic signal having a long wavelength is reproduced. In the above example, the reproduction time length is doubled, so that the wavelength is doubled, and the frequency is conversely converted to half (to a lower sound). The time length correcting means 31 acts to return only the time length of the extended discrete sound signal to the original length without changing the pitch using, for example, the PICOLA method. In the case of a sound signal having periodicity such as voice or musical sound, in the PICOLA method, two adjacent periodic waveforms are added using a tapered window, and the time length of exactly one wavelength is reduced. As described above, in the PICOLA method, since processing is performed in units of periodic waveforms, only the time length can be reduced without changing the pitch. In addition, when the sound signal has no periodicity, a signal having a time length equivalent to a period substantially equal to the time when there is a periodicity is used.
The addition is performed using a tapered window over two adjacent signal sections, and the time length of about one cycle is reduced. In the case of C = 0.5, the output of the acoustic signal interpolation means 21 is subjected to the above-mentioned time length correction, the time length is shortened to し て and returned to the original time length. , And an acoustic signal whose pitch has been converted is extracted from the output terminal 201.

The effect of the hearing aid with a pitch conversion adjusting function that operates as described above will be described with reference to FIG.
FIG. 7 shows 50 words (A) of simple three-syllable words (nouns) for evaluation, such as "head, strawberry, rabbit ..." and "blue,
The three-syllable words for evaluation (B), including nouns, verbs, and adjectives, such as "Speak, move ...", which are somewhat unfamiliar, are subjected to various pitch conversion processing, and the results of 11 elderly hearing-impaired persons who performed hearing evaluation is there. The horizontal axis is the three syllable words for evaluation (A)
7A and 7B show the average pitch frequency of each of the 50 words. In FIG. 7, the position (frequency) indicated by the arrow as Normal is the average pitch frequency of the original sound (without pitch conversion processing) of each evaluation sound,
The plot on the left is the average pitch frequency of the pitch-evaluated evaluation voice, and the plot on the right is the average pitch frequency of the pitch-evaluated evaluation voice. The vertical axis in FIG. 7 indicates a value obtained by averaging the deterioration of the word correct answer rate for all the subjects. Deterioration of the correct answer rate for each subject can be calculated from the word correct answer rate of the average pitch frequency (pitch) at which the highest correct answer rate is obtained for each subject, and other pitch frequencies (pitch)
Is defined as the value obtained by subtracting the word correct answer rate in.

In the case of the three-syllable word (B) for evaluation, the average pitch frequency of the original voice (Normal) is 332 Hz, which is a very high pitched voice. As is clear from FIG. 7, in the case of such a voice, when the pitch is lowered by 1/2 [octave] and the average pitch frequency is converted to 235 Hz, a remarkable effect that the word correct answer rate is improved by about 25% is obtained. I understand.

As described above, according to the first embodiment of the present invention, the pitch conversion rate adjusting means 41 can adjust the input sound signal to the pitch that is most audible, so that the word intelligibility is significantly improved. In addition to the advantage that the sound signal model is not used, there is an advantage that no unpleasant abnormal sound is generated even when the sound contains a musical sound or noise, and an excellent hearing aid effect is obtained.

(Embodiment 2) FIG. 2 shows a configuration of a hearing aid according to a second embodiment of the present invention. In FIG.
Reference numeral 42 denotes a pitch conversion rate adjusting means provided with a manual operation means 62. Other components are the same as those in the first embodiment, and a duplicate description will be omitted.

The second embodiment is configured as described above, and has a feature that the user can freely adjust the pitch by the manual operation means 62. Manual operation means 62
Is realized using a variable resistor or a digital switch. The pitch conversion rate adjusting means 42 operates to output a pitch conversion coefficient (C) to the acoustic signal interpolation means 21 and the time length correction means 31 according to the state of the manual operation means 62. The effect of the hearing aid with a pitch conversion adjusting function that operates as described above will be described with reference to FIG.

As apparent from FIG. 7, in the three-syllable word (A) for evaluation, the average pitch frequency (22
Even if the pitch is lower than 1 Hz), the word correct answer rate does not improve. On the other hand, in the case of a three-syllable word (B) with a high pitched voice, word intelligibility is significantly improved when the pitch is lower than the average pitch frequency (332 Hz) of the original voice, but 235 Hz.
If the voice is converted to a lower voice, the word correct answer rate will decrease. That is, it is desirable that the pitch conversion rate can be adjusted according to the speaker's voice.

As described above, according to the second embodiment of the present invention,
Since the user can adjust the pitch conversion rate according to the speaker's voice, there is an advantage that the user can adjust the pitch conversion rate so as to obtain the maximum hearing aid effect.

(Embodiment 3) FIG. 3 shows a configuration of a hearing aid according to a third embodiment of the present invention. In FIG.
43 is a pitch conversion rate adjusting means for automatically calculating a pitch conversion rate and outputting a pitch conversion coefficient C (t) as a function of time. Reference numeral 303 denotes a signal path for inputting a discrete acoustic signal to the pitch conversion rate adjusting means 43. Other components are the same as those in the first embodiment, and a duplicate description will be omitted.

FIG. 4 is an internal configuration diagram of the pitch conversion means 43 according to the third embodiment. In FIG. 4, reference numeral 303 denotes a signal path for inputting a discrete sound signal, 4301 denotes a means for extracting a pitch frequency from the discrete sound signal by using an autocorrelation method or the like, and 4302 denotes an average for calculating a long-term average value of the pitch frequency. means, 4303 to determine the pitch conversion from the ratio of the higher becomes the pitch frequency is most word correct rate and pitch frequency which is averaged, the conversion coefficient operation means for outputting a transform coefficient C _o, the 4304 current transform coefficients _Is a delay control means for gradually changing the conversion coefficient from the target to the target conversion coefficient Co. For example, when the current conversion coefficient is 1, the delay control unit 4304 outputs the conversion coefficient to the acoustic signal interpolation unit 21 and the time length correction unit 31 as a time function C (t) as shown in FIG.

The operation of the hearing aid having the above-described pitch conversion adjusting function will be described below. The pitch extracting means 4301 calculates an average intra-frame pitch frequency for each frame length of about 20 to 30 ms. The averaging means 4302 performs averaging processing over a long period of time on the frame determined to have periodicity in the frame. By averaging, it is possible to estimate the average voice pitch of a speaker, adjust the pitch so that the speaker's voice can be heard best, and furthermore, adjust the speaker's And the intonation can be kept as it is. Conversion coefficient calculation section 4303 calculates the pitch conversion coefficient C _o from the ratio of the average voice pitch speakers (average pitch frequency) as the most easily audible pitch frequency shown in FIG. The delay control unit 4304 acts to prevent a sudden change in the pitch when the cycle of the input acoustic signal is greatly changed instantaneously, thereby preventing a sense of incongruity. For example, even when the speaker changes and the pitch of the voice changes greatly, it acts so as to gradually change to the target pitch conversion rate.

When the current pitch conversion coefficient is 1, the delay control means 4304 sets a as a time constant, as shown in equation (2).
Output the transform coefficient C (t) as a function of time. C (t) = _Co- (1- _Co ) exp (-a) (2)

As described above, according to the third embodiment of the present invention, the average pitch frequency of a speaker can be estimated and automatically adjusted to the pitch that is most audible, and the pitch conversion rate due to a change in the speaker can be obtained. Since the rapid fluctuation of the pitch can be reduced, there is an advantage that the pitch conversion with less discomfort can be automatically performed.

(Embodiment 4) FIG. 5 shows an internal configuration of a pitch conversion rate adjusting means 43A according to a fourth embodiment of the present invention. The entire configuration is the same as that of the third embodiment shown in FIG. In FIG. 5, reference numeral 4305 denotes voice quality change restricting means, and the other components are the same as those in the configuration of FIG.

Referring to FIG. 5, voice quality change limiting means 4305
When the pitch conversion rate exceeds a certain value (for example, 〔[octave]), it acts to limit the pitch conversion rate to less than that value. By changing the pitch, the voice quality changes at the same time as the pitch of the voice changes. Voice quality is peculiar to each person, and it is usually possible to judge who is just listening to the voice. If the voice quality changes too much due to the pitch conversion, it becomes difficult to determine who the voice is. Therefore, in the fourth embodiment, the pitch conversion rate is limited so as not to become too large.

As described above, according to the fourth embodiment of the present invention, since the limitation is made so that excessive pitch conversion is not performed, the natural voice quality of the speaker is not greatly changed. This has the advantage that the pitch can be easily converted.

(Embodiment 5) FIG. 6 shows an internal configuration of a pitch conversion rate adjusting means 43B according to a fifth embodiment of the present invention. The entire configuration is the same as that of the third embodiment shown in FIG. 6, reference numeral 4306 denotes a priority switching unit, 65 denotes a manual operation unit, and other components are the same as those in FIG.

In FIG. 6, priority switching means 4306
_Operates such that the adjustment of the conversion coefficient _Co by manual operation is automatically given priority by operating the manual operation means 65. Acoustic signals input from the microphone of the hearing aid include various types of sounds, and not all sounds desired to be heard (for example, sounds) are necessarily input. Therefore, the hearing aid effect by the pitch conversion is not always effective in all cases. In the fifth embodiment, in addition to the means for automatically adjusting the pitch so as to obtain the highest hearing aid effect,
At the discretion of the user, the pitch can be manually adjusted with priority.

As described above, according to the fifth embodiment of the present invention, the pitch adjustment by manual operation is prioritized according to the user's judgment. Has the advantage that the user can select the pitch adjustment means with the greatest hearing aid effect

[0035]

As described above, according to the present invention, the pitch can be adjusted so that the highest hearing aid effect can be obtained automatically or manually by the user without depending on the input acoustic signal. Thus, the rate of correct answering of words can be significantly improved, and an excellent hearing aid effect with less discomfort due to abnormal noise or voice quality change can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a hearing aid according to a first embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a hearing aid according to a second embodiment of the present invention.

FIG. 3 is a block diagram showing a configuration of a hearing aid according to a third embodiment of the present invention.

FIG. 4 is a block diagram showing an internal configuration of a pitch conversion rate adjusting unit according to a third embodiment.

FIG. 5 is a block diagram showing an internal configuration of a pitch conversion rate adjusting unit according to a fourth embodiment.

FIG. 6 is a block diagram showing an internal configuration of a pitch conversion rate adjusting unit according to a fifth embodiment.

FIG. 7 is a characteristic diagram showing a deterioration characteristic of a correct answer rate of a pitch-converted word voice heard by an elderly hearing-impaired person.

FIG. 8 is a characteristic diagram showing a time change of a pitch conversion coefficient C (t).

FIG. 9 is a block diagram showing a configuration of a hearing aid in a conventional example.

[Explanation of symbols]

REFERENCE SIGNS LIST 1 AD conversion unit 2 PARCOR analysis unit 3 PARCOR synthesis unit 4 control unit 5 DA conversion unit 10 audio signal input terminal 11 AD conversion unit 20 output terminal 21 audio signal interpolation unit 31 time length correction unit 41, 42, 43, 43 A, 43 B Pitch conversion rate adjusting means 51 DA conversion means 62, 65 manual operation means 101 sound signal input terminal 201 terminal for outputting a pitch conversion signal 303 signal path for inputting a discrete sound signal 4301 pitch extraction means 4302 averaging means 4303 conversion coefficient calculation Means 4304 Delay control means 4305 Voice quality change limiting means 4306 Priority switching means

────────────────────────────────────────────────── ─── Continuation of front page (56) References JP-A-5-316597 (JP, A) JP-A-6-140856 (JP, A) JP-A-9-46795 (JP, A) (58) Field (Int.Cl. ⁷ , DB name) H04R 25/00 H03G 11/00

Claims (5)

(57) [Claims] 1. A means for extending a time length by interpolating a discrete sound signal, thinning out a periodic waveform in synchronization with a cycle in a periodic sound signal section, and removing a periodic waveform in a non-periodic sound signal section. A hearing aid with a pitch adjustment function comprising: means for correcting a time length by thinning out a signal having a cycle length substantially equal to a certain signal section; and pitch conversion rate adjusting means for determining a time length expansion rate of a discrete sound signal. . 2. The hearing aid with a pitch adjusting function according to claim 1, wherein the pitch conversion rate adjusting means includes means by which a user of the hearing aid can adjust a pitch by manual operation. 3. A pitch adjusting function according to claim 1, wherein the pitch conversion rate adjusting means includes means for automatically adjusting a pitch so that the voice is most easily heard. hearing aid. 4. The pitch conversion rate adjusting means automatically adjusts the pitch so that the voice pitch becomes the most audible, and automatically adjusts the pitch conversion rate so that the voice quality does not become unnatural. The hearing aid with a pitch adjusting function according to claim 1, further comprising a limiting unit. 5. The pitch adjustment according to claim 4, wherein the pitch conversion rate adjusting means can adjust the pitch conversion rate by a user's manual operation prior to the automatic adjusting means. Hearing aid with function.