JP2000059893A - Hearing aid device and its method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow an audibly handicapped person to discriminate voice signals heard in plural directions. SOLUTION: Selector switches 105, 106 select transfer functions 107-109 so that a sound signal from a sound source to which an audibly handicapped person opposes is heard by the audibly handicapped person in the directly opposite position and the voice signal from the other sound source is heard with a relative position among sound signals of sound sources place in plural directions. The selector switches 105, 106 decide a transfer function by a detection signal of a head direction detector 111.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an audio-listening assisting apparatus and method for assisting a person with hearing impairment in listening to audio, and more particularly, to emphasizing the localization of a sound source. The present invention relates to an audio listening assisting device and method.

[0002]

2. Description of the Related Art Conventionally, in a television viewing system for the elderly with hearing impairment, a television sound is directly input from a television receiver to a hearing aid using an FM and an induction coil, and the television sound is listened. Auxiliary system (Reference 1, Hiroshi Kondo, Kazuhiko Hikita “Wireless hearing aid system” Nati
onal Technical Report Vol. 42 No. 2 Apr 1996 pp 88-
95). In this method, the TV sound can be heard without being affected by the surrounding environment without amplifying the surrounding noise at the same time. In some cases, an earphone-integrated microphone is used to switch between a TV sound by FM and a conversation sound with the surroundings, or to select a mixed sound. In these systems, it is only possible to listen to either the television sound or the conversation sound, or to simply select a mixed sound. Also, the selection of the main voice is manually switched.

On the other hand, as a hearing aid system for assisting selective listening, a binaural hearing aid system has been considered. According to this method, the loudness of sound presented simultaneously to the left and right ears (loudness) is increased by hearing aid to both ears compared to the case where the sound is presented to one ear (loudness addition).
To ensure the required loudness without presenting loud sounds close to the discomfort threshold, and to use the sound image localization ability left by the user to separate and listen to the required sounds to improve the selective listening ability. It aims at improvement (masking release effect). However, the improvement of selective listening ability such as sound image localization by binaural hearing aid varies from person to person and may not be seen as being improved (Ref. 2, Osamu Wada, Kenji Ozawa et al. Considerations on Binaural Wear "IEICE Technical Report SP95-138 (1996-3) pp 23-30, Reference 3, Monica L Hawley and H. Steven Colburn" Speech In
telligibility and Localization in a Complex Enviro
nment Relationship to Binaural Hearing "Hearing Ai
d Research and Development Conference September 22
-24 1007, http://www.nih.gov/nidcd/hard/abstrcts/h
awley.htm).

[0004] In addition, in a case where a plurality of sound sources having different directions of arrival exist, it is difficult for a hearing-impaired person to hear the target sound while the number of disturbing sound sources increases and the hearing of the target sound becomes difficult. It is reported that the degree increases (Ref. 4, Joan M. Besing and
Lanet Koehnke "Speech Recognition in Listeners wit
h Impaired Hearing: Effects of Listening of Listeni
ng of Paradigm "Hearing Aid Research and Development
Conference September 22-24 1997, http: //www.nih.go
v / nidcd / hard / abstrcts / besing.htm). Therefore, if the number of interfering sound sources increases and it becomes necessary to have not only the ability to localize sound images but also the ability to decompose sound images, a binaural hearing aid that is simply mixed and amplified may not provide sufficient effects. There is.

[0005] In the aging society that is about to be welcomed, it has been reported that the degree of disability may be greater in men than in women due to gender differences in the tendency of hearing impairment (Ref. 5, Jay). D. Pearson, Christopher H.
Morrell. Et al "Gender difference ina longitudinal
study of age-associated hearing loss "J.Acoust.So
c.Am 97 (2) February 1995 pp 1196-1205), "Grandpa" with disability and "Grandma" with relatively good hearing
It is expected that there will be more opportunities to watch TV together.

[0006]

In such an environment, when a person with a hearing impairment uses the above-mentioned audio-listening assistance device, there are the following problems. When a device for manually switching between the television sound and the family member's sound by the selection switch is used, for example, a person with hearing impairment can hear only one of the television sound and the family member's sound. For this reason, when a person with a hearing impairment is listening to the television sound, even if a family member speaks, the person does not notice.

[0007] In addition to the above-described two audio modes of television and family members, there is an apparatus which can select a mode for mixing and amplifying television audio and family audio. There is a problem that making the selection is troublesome for a user of the apparatus (a person with hearing impairment).

Another problem is that it is difficult for a person with a hearing impairment to discriminate a desired sound from the above two sounds even if the TV sound and the sound of the family are simply mixed and amplified. is there.

Accordingly, an object of the present invention is to enable a user of a device to hear sounds generated from a plurality of sound sources, and to discriminate (recognize) specific sounds among the plurality of sound sources. It is an object of the present invention to provide a device and method for assisting voice listening which can be improved.

[0010]

In order to achieve such an object, to achieve such an object, claim 1 is provided.
The invention is characterized in that a plurality of sound input means for inputting sounds generated from a plurality of sound sources in the form of sound signals, a selecting means for selecting one sound source among the plurality of sound sources, Having a first localization position so that the sound of the sound source selected by the selecting means can be heard from the front,
The localization positions of the sounds of the plurality of sound sources input from the plurality of sound input units are controlled such that the sounds of the other sound sources have a second localized position corresponding to the relative positional relationship between the plurality of sound sources. Control means.

According to a second aspect of the present invention, in the audio-listening assisting device according to the first aspect, the selecting means selects the plurality of sound sources according to a direction facing the means.

According to a third aspect of the present invention, there is provided the audio-listening assistance device according to the first aspect, further comprising means for removing a redundant component from the audio signal output by the audio input means.

According to a fourth aspect of the present invention, in the audio-listening assisting device according to the third aspect, the redundant component is generated when the audio is reproduced.
It is characterized by being a signal component giving a spatial spread.

According to a fifth aspect of the present invention, in the audio-listening assisting device according to the third aspect, one of the plurality of sound sources is a television receiver, and the means for removing the redundancy is a television receiver. The present invention is characterized in that a stereo video signal of a receiver is converted into a monaural video signal, thereby removing a signal component giving a spatial spread.

According to a sixth aspect of the present invention, in the audio-listening auxiliary device according to the third aspect, the means for removing the redundancy comprises:
It is characterized in that signal components other than a specific frequency band unique to human speech are removed as reverberation components.

According to a seventh aspect of the present invention, in the audio-listening assisting device according to the third aspect, the means for removing the redundancy includes:
It is characterized in that signal components in a specific frequency band unique to human voice are emphasized.

According to the present invention, voices generated from a plurality of sound sources are input in the form of voice signals, one of the plurality of sound sources is selected, and the selection of the plurality of sound sources is performed. The first sound that can be heard from the front of the sound source
And controls the localization positions of the input sounds of the plurality of sound sources so that the sounds of the other sound sources have the second localization positions corresponding to the relative positional relationships of the plurality of sound sources. It is characterized by the following.

According to a ninth aspect of the present invention, in the voice listening assisting method according to the eighth aspect, the plurality of sound sources are selected by detecting a direction of a human head to be subjected to hearing assistance. I do.

According to a tenth aspect of the present invention, in the audio listening assisting method according to the eighth aspect, a redundant component is removed from the audio signal.

[0020] According to an eleventh aspect of the present invention, in the audio listening assisting method according to the tenth aspect, the redundant component is a signal component that gives a spatial spread during audio reproduction.

According to a twelfth aspect of the present invention, in the audio listening assisting method according to the tenth aspect, one of the plurality of sound sources is a television receiver, and the stereo video signal of the television receiver is converted into a monaural signal. It is characterized in that a redundant component giving a spatial spread is removed by converting into a video signal.

According to a thirteenth aspect of the present invention, in the audio-listening assisting method according to the tenth aspect, reverberation components other than a specific frequency band unique to human speech are removed as redundant components.

According to a fourteenth aspect of the present invention, in the voice listening assisting method according to the tenth aspect, a signal component in a specific frequency band unique to human voice is emphasized.

[0024]

Embodiments of the present invention will be described below in detail with reference to the drawings.

In the present embodiment, in order to assist the elderly in selective listening, the sound environment at the time of watching TV is divided into two sounds, a TV sound and a conversation sound with the family, and the contents of the sound language included in each sound are understood. Remove redundant components that do not contribute to
Separate / relocate and present.

The removal of the redundant components of the two voices is performed, for example, by removing the spatial spread of each voice. TV sound is converted from stereo to monaural, and conversational sound is removed from reverberation components.

Separation / rearrangement presentation of two voices is performed by controlling the sound image localization directions of the two voices in earphone listening with both ears, for example. When the user selects the TV sound as the main sound, control is performed such that the localization direction of the TV sound is set to the front TV direction, and the conversation sound of the family is positioned to the rear right (left).

Also, when the user selects the conversation with the family as the main sound, the television sound is localized in the backward left (right) direction so that the conversation with the family is localized in the direction of the front and the conversation partner. To control. For control of the sound image localization direction, convolution of a head related transfer function, which is an existing technique, is used.

In the control of the sound image localization direction, the direction of the head of the listener is detected to automatically determine which of the television sound and the human sound has been selected as the main sound by the listener. . Also, control is performed so that the selected main sound can be heard from the front so that the listener can hear the main sound better than the sub sound.

A method of assisting voice listening for executing such voice control will be described with reference to FIG. FIG. 1 shows three head-related transfer functions used in a later-described audio-listening assist device under an assumed television viewing environment.

In FIG. 1, reference numeral 108 denotes a head-related transfer function prepared for localizing a sound source in front of the front. 109
Is a head-related transfer function prepared for localizing the sound source to the right rear. Reference numeral 107 denotes a head-related transfer function prepared for localizing the sound source to the left rear.

The head-related transfer function 108 is an impulse response obtained by placing a sound source in front of the user and measuring impulse signals emitted from the sound source at the entrances of the external auditory canals of both ears of the user. Similarly, the head-related transfer functions 109 and 107 are impulse responses measured at unknown entrances of the outer ears of both ears of the user when the sound source is placed on the right rear and left rear.

By convolving the selected transfer function among these transfer functions into an audio signal with a small clue of spatial spread or near / far position, the user can determine the localization direction of the sound source in front front, right rear, The virtual localization direction of the sound source is controlled so that the sound source can be perceived as being at the rear left. In addition, in each of the localization directions, a transfer function is selected so as to be localized at a distant position so that a user having a reduced selective listening ability can sufficiently select and listen with the remaining selective listening ability.

FIG. 2 shows the configuration of an audio-listening auxiliary device used in an assumed television viewing environment. Although FIG. 2 will be described later, FIG. 2 shows a case where the device user (a person with hearing impairment) selects television sound as the main sound by directly facing the television receiver.

In FIG. 2, reference numeral 100 denotes a T which incorporates the above-mentioned voice emphasizing device and reproduces and outputs (first) voice and video.
V receiver. An audio component of a specific frequency band is enhanced by the audio enhancement device, and a signal (referred to as an audio signal) in which a component contributing to a sense of spatial expansion and direction perception is removed from the TV video signal by the audio enhancement device is received by the TV receiver. The signal is input from the machine (100) to the convolution processor (102). In the case of this example, the video signal output unit (not shown) in the TV receiver (100) operates as one of the plurality of audio input units.

Reference numeral 101 denotes a sound collecting device such as a microphone that collects the second sound for conversation with the family and outputs the sound in the form of an electric signal (sound signal), and includes a sound emphasizing device. The voice emphasizing device emphasizes the signal component of a frequency band peculiar to the voice in the voice signal collected by the voice collecting device (101), and also has a sense of spatial expansion and a signal component contributing in the vicinity of a direction, and room reverberation. The component is removed from the collected audio signal. Further, signal components other than the frequency band unique to the sound may be removed. The sound collection device (101) constitutes a plurality of audio input units together with the video signal output unit.

Reference numeral 102 denotes a transfer function convolution calculator for controlling the sound image localization direction of the TV sound.

Reference numeral 103 denotes a transfer function convolution arithmetic unit for controlling the localization direction of the speech sound image.

Reference numeral 104 denotes a mixing amplifier that mixes and amplifies a TV voice and a conversation voice whose sound image localization direction is controlled.

Reference numeral 105 denotes a selection switch for selecting a transfer function to be convolved with the TV sound.

Reference numeral 106 denotes a selection switch for selecting a transfer function to be convolved with the conversation voice.

Reference numeral 107 denotes a holding circuit for holding a transfer function for localizing a sound image (TV sound and conversation sound in this example) to the rear left.

Reference numeral 108 denotes a holding circuit for holding a transfer function for localizing the sound image in front of the front.

Reference numeral 109 denotes a holding circuit for holding a transfer function for localizing the sound image to the right rear.

The transfer function convolution calculators 102, 10
For 3, a multi-stage filter is used, and the transfer function can be given to the transfer function convolution unit in the form of a filter coefficient.

Reference numeral 110 denotes a voice presentation device such as headphones.

Reference numeral 111 denotes a head orientation detector for detecting the orientation of the user's head. The head orientation detector 111 detects the orientation of the detector itself as the head orientation, and outputs a measurement result in the form of an electric signal. An electronic azimuth instrument can be used.
The detector 111 may be installed in headphones or the like. In the example of FIG. 2, the head orientation detector (111) is a TV receiver (10
0), the device user selects a transfer function that allows the user to hear the TV sound from the direction directly facing the user and a transfer function that allows the sound of the family to be heard from the right rear.
W ((105, 106) is selected.

On the other hand, in the case where the head orientation detector 111 detects the orientation toward a family member, a description will be given with reference to FIG. The selection switches (105, 106) select a transfer function that can be heard from the direction in which the voice is directly facing.

In FIG. 2, single and double lines in the figure indicate monaural and stereo.

In the present embodiment, the sound of the TV receiver (100) is subjected to processing such as stereo-monaural conversion, enhancement of the speech sound band, and removal of reverberation by a time-base autocorrelation filter in the sound emphasizing device. The component that is perceived to be spread is removed and output.

Here, in the stereo-monaural conversion, when the television program audio is a stereo broadcast, only the signals having a strong correlation between Lch and Rch are output, so that important sounds in the program (voices of an announcer such as news, Narration, etc.), and suppresses SE and BG components that are perceived as spreading the sound image. In order to construct this portion at low cost, the (L + R) signal may be used as a monaural signal.

The enhancement of the speech voice band contributes to the perception of the spatial spread of the sound source, but the frequency band not involved in the perception of the speech voice is reduced by a voice emphasis device (filter, etc.).

Further, since the echo component of the uttered voice also gives a perception of the spatial spread of the sound source and is considered to be a redundant component, the filtering process of the voice emphasizing device for reducing the correlation on the time axis is performed. Echo and reverb components can be removed by applying an appropriate window width.

The head-related transfer function (108) designed to be located in front of the front of the TV receiver output signal from which the components contributing to the sense of spatial expanse and direction perception have been removed is convolved with the convolution unit (102). ), And becomes a stereo signal of L and R in which the sound image is controlled forward and forward.

On the other hand, as for the uttered voice of the family, only the uttered voice of the family sitting at the same place can be collected at all times by using a directional microphone or a microphone array sound collector (101). Further, the collected uttered voice is removed by an adaptive voice emphasis device (filter) that removes a temporal correlation in the same manner as the output signal of the TV receiver, thereby removing spatial spread such as room reverberation. And the sound collector output (l
01).

As described above, in the case of FIG. 2,
Similarly to the output signal of the TV receiver, the output of the sound collector from which the components contributing to the spatial feeling and direction perception are removed is (10
In 1), a convolution calculator (103) convolves a transfer function (109) designed to be localized to the right rear as viewed from the user, so that a stereo image of L and R whose sound image direction is controlled to the right rear. Signal.

The television sound and the conversation sound whose sound image localization directions are respectively controlled are synthesized by the mixing amplifier (104) and presented to the user's both ears by a sound presentation device (1110) such as an earphone.

In the present embodiment, the transfer function selection SW (10
5, 106) is controlled by a head orientation detector (11l) incorporated in the audio presentation device (1110). When the user is facing the television, the control is performed so that the television sound is localized in front of the front, and the uttered voice of the family is localized in the right rear. On the other hand, when the user is facing a direction other than the television direction (the direction of the family), the TV sound is localized to the rear left, and the uttered voice of the family is localized to the front front.

FIG. 3 shows a case where the user is facing a direction other than the television direction (direction toward a family member) under the same television viewing environment as in FIG. 3, the same parts as those in FIG. 2 are denoted by the same reference numerals,
Detailed description is omitted. FIG. 3 shows a case where the apparatus user can hear the conversation voice in front of the user and has selected the conversation voice as the main voice.

In the case of FIG. 3, the head orientation detector (11
(Transfer function) selection SW whose connection destination is controlled by 1)
In (105, 106), since the user has selected the conversation voice (facing in a direction other than the television), the selection SW
(105) selects the transfer function (107) by selecting SW (10).
6) selects the transfer function (108). Therefore, the conversation sound is controlled to be localized in front of the user, and the television sound is controlled to be localized in the left rear of the user.

As described above, in the above-described embodiment, when the user of the audio-listening assist device faces a sound source in an arbitrary direction, the sound output from the sound source starts from the direction facing the user of the device. The sound signal presented to the user of the device is audible, and the sound output from the other sound source is heard from a direction corresponding to the relative positional relationship between the facing sound source and the other sound source. Control the localization position. This allows the device user to recognize the sound of the sound source facing himself as the main sound,
Sounds from sound sources in other directions can be simultaneously recognized as sub-sounds. Further, since the sounds from the plurality of sound sources have directionality, even when presented at the same time, the apparatus user can clearly discriminate (distinguish) these sounds.

The following embodiment can be carried out in addition to the above-described embodiment.

1) The head orientation detector is not limited to the electronic compass, and other measuring instruments can be used. Also, the head orientation detector does not need to be installed on the head, and may be installed on the chest, for example. If it is not necessary to automatically detect the direction, for example, an instruction means for instructing the direction by a dial operation of the device user may be used.

2) As the voice input means, a circuit for converting air vibration into an electric signal, such as a microphone, or a circuit for inputting the electric signal itself can be used.

3) As a sound source to be subjected to the apparatus, electric equipment which reproduces and outputs sound, such as human voice, so-called sound, a television receiver, a radio receiver, and a personal computer can be used.

4) In the above embodiment, the transfer function is set in advance. This is suitable when the positions of a plurality of sound sources are fixed in advance. However, an arbitrary position and number of sound sources are designated by well-known instruction means (for example, numeric keypad), and a transfer function corresponding to the instruction is designated. Can be used selectively. In order to select such an arbitrary number of transfer functions, a programmable logic array or the like may be used as a selection switch.

5) In order to reproduce and present the voices of the family members more accurately, the frequency band of the voice of the family member (s) may be measured in advance and reflected on the apparatus. Depending on the application, the frequency band of the voice of the unspecified speaker may be used for voice emphasis, or the specific speaker and the unspecified speaker may be switched.

[0068]

As described above, according to the present invention, in a future aging society, it is possible to improve the television viewing environment of elderly people who have difficulty in selective listening due to deterioration of various auditory characteristics. More specifically, claims 1, 2,
According to the eighth and ninth inventions, when a plurality of different audio signals are reproduced, each of the audio signals has a localization position,
Also, an audio signal that can be heard from the direction facing the user of the device is treated as the main audio. For this reason, it is possible to accurately detect the sound of the sound source that the user wants to hear, and identify the sound by the localization position. Also, by automatically detecting the orientation of the device user's head, etc.,
Since the sound source of the main sound (the sound source in the facing direction) can be selected, the user of the device is released from manual operation.

According to the third to seventh and tenth to fourteenth aspects of the present invention, acoustic components other than speech are removed as redundant components.
Alternatively, since the audio component is emphasized, when the audio signal is reproduced, the audio content can be easily and accurately grasped.

[Brief description of the drawings]

FIG. 1 is a configuration diagram for explaining a selective listening assisting method according to the present invention.

FIG. 2 is an overall configuration diagram showing a configuration of a voice listening assistance device according to the present invention.

FIG. 3 is an overall configuration diagram for explaining an operation of the audio listening assist device according to the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 100 television receiver 101 sound collector (output) 102, 103, 104 convolution calculator 107, 108, 109 (head) transfer function 110 voice presentation device 111 head orientation detector

Claims (14)

[Claims] A plurality of sound input means for inputting sounds generated from a plurality of sound sources in the form of sound signals; a selecting means for selecting one of the plurality of sound sources; and a plurality of sound sources. Has a first localization position so that the sound of the sound source selected by the selection means can be heard from the front, and the sound of the other sound sources corresponds to the relative positional relationship between the plurality of sound sources. Control means for controlling the localization positions of the sounds of the plurality of sound sources input from the plurality of sound input means so as to have the localization positions. 2. The audio-listening assisting device according to claim 1, wherein the selecting means selects the plurality of sound sources according to a direction facing the selecting device. 3. The audio-listening assistance device according to claim 1, further comprising a unit for removing a redundant component from the audio signal output by the audio input unit. 4. The audio-listening assistance device according to claim 3, wherein the redundant component is a signal component that gives a spatial spread during audio reproduction. 5. The audio-listening assistance device according to claim 3, wherein one of the plurality of sound sources is a television receiver, and the means for removing the redundancy includes a stereo image of the television receiver. An audio-listening assisting device, which converts a signal into a monaural video signal to remove a signal component giving a spatial spread. 6. The audio-listening assistance device according to claim 3, wherein the means for removing the redundancy removes a signal component other than a specific frequency band unique to human speech as a reverberation component. Auxiliary audio listening device. 7. The audio-listening assistance device according to claim 3, wherein the means for removing the redundancy emphasizes a signal component in a specific frequency band unique to human speech. apparatus. 8. A sound generated from a plurality of sound sources is input in the form of a sound signal, one sound source is selected from the plurality of sound sources, and the sound of the selected sound source is selected from the plurality of sound sources. Have a first localization position such that they can be heard from the front, and the sound sources of the other sound sources have the second localization position corresponding to the relative positional relationship between the plurality of sound sources. A method for assisting in listening to a voice, comprising controlling a localization position of the voice. 9. A method according to claim 8, wherein said plurality of sound sources are selected by detecting a direction of a human head to be subjected to hearing assistance. . 10. The audio-listening assistance method according to claim 8, wherein a redundant component is removed from the audio signal. 11. The audio listening assisting method according to claim 10, wherein the redundant component is a signal component that gives a spatial spread during audio playback. 12. The audio-listening assisting method according to claim 10, wherein one of the plurality of sound sources is a television receiver, and converts a stereo video signal of the television receiver into a monaural video signal. A method for assisting voice listening, characterized by removing redundant components that give a spatial spread. 13. A method according to claim 10, wherein a reverberation component other than a specific frequency band specific to a human voice is removed as a redundant component. 14. A method according to claim 10, wherein a signal component in a specific frequency band specific to human voice is emphasized.