CN114040308B - Skin hearing aid device based on emotion gain - Google Patents

Abstract

A skin-acoustic hearing aid device based on emotional gain, belonging to the field of hearing aid technology, including a microphone, a hearing aid host and electrode stickers, wherein the microphone can be set independently and connected to the hearing aid host through wires, or can be set on the hearing aid On the host, the electrode stickers use composite planar electrodes, including one or two in parallel, pasted on the skin surface behind the human ear, connected to the hearing aid host through wires, and the hearing aid host has a built-in power supply. The present invention has a novel structure and ingenious conception, and uses the emotional characteristics recognized by the computer to increase the skin hearing effect, so that the hearing aid effect is better, the ability to receive sound frequencies is strong, and at the same time, it effectively reduces the learning difficulty and threshold for hearing-impaired people. The effect is remarkable.

Description

A skin-acoustic hearing aid based on emotional gain

technical field

The invention relates to the technical field of hearing aids, in particular to skin hearing aids, in particular to a skin hearing aid technology and application based on emotional gain.

Background technique

According to the 2018 report of the World Health Organization, there are 360 million hearing-impaired people in the world, accounting for 5.3% of the total global population. In my country, at present, there are as many as 27.8 million hearing-impaired people, accounting for 1.679% of the national population, and there are about 30,000 newly born disabled people due to various reasons every year.

Hearing impairment refers to the organic or functional abnormalities of the nerve centers at all levels of sound transmission, sensory sound and comprehensive analysis of sound in the auditory system, resulting in varying degrees of hearing loss, commonly known as deafness. In fact, only severe hearing loss is called deafness, which is manifested by the patient's inability to hear any speech in both ears, and those whose hearing loss has not reached this level of severity are called hearing loss. After the human body grows and develops, the hearing begins to lose irreversibly year by year. In this process, the speed of hearing loss may be accelerated due to the influence of some external factors, such as drugs, noise environment, trauma, etc. Most adults have Unavoidable in the process is the rapid hearing loss due to answering calls for a long time, wearing earphones/earplugs for a long time, and being in a noisy environment for a long time. At present, there is no effective treatment for hearing impairment. Restoration and improvement methods, which make the population of mild and moderate hearing-impaired people in our country grow at an extremely fast rate, and far exceed the statistical quantity.

In order to improve this situation, on the one hand, actively cultivate people's good listening habits and slow down the speed of hearing loss, and on the other hand, use hearing aids to compensate for the hearing loss of people with severe or above hearing impairment. For the latter, the existing hearing aid devices include hearing aids and cochlear implants. Hearing aids can compensate for the hearing loss of people with severe hearing impairment and below, and are relatively cheap. There are often noises that are difficult to eliminate, which will further cause the user's hearing loss while compensating for the user's hearing loss. At the same time, the use of deaf patients is completely ineffective; cochlear implants are achieved through surgery, which can improve the patient's hearing to completely normal state, but the price is expensive, and the success rate of the operation is not high. The operation will completely destroy and replace the patient's original hearing organ. No matter whether it is successful or not, the patient's hearing organ cannot be repaired. Hearing is irreparable. Therefore, it is imperative to study better hearing aid solutions, and skin hearing technology is one of the excellent ideas.

Studies have shown that human skin can sensitively feel various stimuli, such as pressure, vibration, electricity, etc., and the nervous system will send these stimuli from the skin to the brain for identification and processing. Since the hearing of the human ear itself is amplified through a series of vibrations—the vibrations are recognized and converted into electrical signals—the nervous system transmits them to the brain—the brain recognizes them, so the external sound can also be completely “heard” through the skin, which is important for part of the sense of hearing. It is great news for the hearing-impaired persons whose organs are destroyed and completely lose their hearing. In the existing patent "Variable Pressure Skin Hearing Device" (Application No. 200410026265.5), a device made by using skin listening technology has been disclosed. It is a kind of hearing aid device, through which the totally deaf can also obtain the hearing of speech and frequently encountered sounds.

However, the environment in which we live is full of various sounds. In addition to the sound of people talking, there are also various background noises. We who have hearing have learned to analyze various sounds through long-term training and learning. Classify various sounds, which are human voices, and which are various environmental noises, and then the brain will focus on the type of sound it wants to hear, and ignore other types of sounds, which makes the sound in our ears The world is full of layers. For a person with a congenital complete loss of hearing, this is a completely unfamiliar field, directly facing all kinds of voices, after losing the efficient learning ability of the "childhood-juvenile" stage, The brain needs a long time to train and learn, so for adults, the technical solution proposed by this patent is not friendly, and the difficulty and threshold of use are very high.

The basis of human-computer emotional interaction comes from computer applications, through algorithms and a lot of learning to simulate the computer's "artificial psychology" and "artificial emotion" capabilities, by analyzing the model characteristics of various emotions to achieve emotional modeling, and further computer Based on the research of domestic and foreign experts, all of these are feasible. On this basis, the Chinese Academy of Sciences will further apply the above theory to the understanding and synthesis of emotional speech, which will The process of human-emotional interaction has taken a big step forward. On the one hand, it can enable computers to accurately determine human emotions by recognizing speech, and on the other hand, it can also convert text into speech or other signals with emotions.

Applying the above ideas to hearing aids for the hearing-impaired can greatly improve the hearing effect of the hearing-impaired, and the effect of compensating their hearing is very obvious.

Contents of the invention

In order to solve the deficiencies of the prior art, the present invention proposes a skin-acoustic hearing aid device based on emotional gain to provide hearing compensation for hearing-impaired persons.

The technical problem to be solved in the present invention is achieved through the following technical solutions:

A skin-acoustic hearing aid device based on emotional gain, including a microphone, electrode stickers, and a hearing aid host connected to the microphone and the electrode stickers. In the present invention, the hearing aid host includes a sound preprocessing module, an emotion analysis module, Emotional additional module, information processing module and sound information output module, wherein the sound preprocessing module is electrically connected with the microphone, including a digital module, the digital module converts the analog signal output by the microphone into a digital signal, and the emotion analysis module is respectively connected with the sound preprocessing module and the The emotional additional module is connected, and the information processing module is respectively connected to the sound preprocessing module, the emotional additional module and the sound information output module, and the information processing module processes the digital signal from the sound preprocessing module to make it convenient for disabled people to learn and train, and Send it to the sound information output module, the sound information output module is respectively connected to the information processing module and the electrode patch, and the sound information output module converts the digital signal into an analog signal, and transmits it to the human skin through the electrode patch.

In the present invention, the information processing module separates the preprocessed digital signal, processes the human voice and the background sound separately, and the human voice is combined with the background sound after being adjusted by the emotional additional module, and sent to the sound information output module. Among them, the information processing module is connected with the emotional additional module, and after compiling the digital signal of the human voice into a digital signal of an easily recognizable text, the syllable and speech rate are adjusted by the emotional additional module, and the output is a digital signal that can be expressed as an analog voice. After re-reading, the dissonant components in the human voice language are cleaned up—such as dialects—and transformed into simulated human voices that are easier to accept and recognize, generally in Mandarin, which is easy to understand and reduces the difficulty and threshold of learning.

Further, the emotional analysis module is connected with an emotional feature database, extracts the emotional features in the digital signal obtained by the digitization module, compares them with the emotional feature data, and determines the emotional features contained in the language represented by the emotional features expressed by the digital signal. The emotion is sent to the emotion additional module, and the corresponding emotion feature is superimposed into the digital signal processed by the information processing module through the emotion additional module.

In the present invention, the sound information output module includes a filter and a booster, wherein the information processing module, the filter, the booster and the electrode stickers are connected in sequence, and there are several sets of filters and boosters Parallel setting.

Further, the filter is a multi-channel band-pass filter, each set of filters selects a different center frequency, and the selection range is 15Hz~15kHz.

Further, there are preferably 48 groups of filters and boosters, and 48 channels of different frequencies are used to output to the electrode stickers.

In the present invention, the electrode stickers are connected to the hearing aid host through wires, and the electrode stickers are pasted on the skin with dense distribution of human nerves.

Further, the position where the electrode paste is pasted is on the surface layer of the skin behind the human ear.

In the present invention, the sound preprocessing module includes a denoising module and a syllable checking module.

In the present invention, the hearing aid host has a built-in power supply.

In the present invention, the electrode paste adopts composite planar electrodes.

In the present invention, the extraction of described emotional features and the establishment of emotional feature database are based on Chinese speech, and some experts and scholars including Beijing University of Aeronautics and Astronautics in my country have created Chinese speech emotion extraction and modeling methods combined with emotional features, and by Based on the established database, the extraction method of the emotional features of Chinese speech is: specifying the emotional feature database specification, including speaker specification, recording script design specification, audio file naming specification, etc.; collecting emotional feature data: emotional feature pleasure, activation degree , Predominance degree (PAD) evaluation, that is, the PAD subjective listening evaluation test is conducted on the emotional feature data by at least ten evaluators who are different from the speaker. The modeling method of Chinese speech emotional features is as follows: firstly, according to the Fisher ratio, select the speech features to train the gender recognition support vector machine model (SVM); secondly, establish emotional feature hidden Markov models (HMM) for male voice and female voice respectively, and use the SVM gender The recognition result selects the corresponding HMM for emotional feature classification.

Compared with the prior art, the present invention has the following advantages:

1) Based on the principle of skin hearing, it ensures that the hearing-impaired can hear the sound, and adjusts it through emotional gain to make the sound heard by the hearing-impaired more real;

2) During the whole processing process, the analog signal output by the microphone is converted into a digital signal. After the processing is completed, the analog signal is converted into an analog signal and output by the electrode sticker, which makes the signal more convenient to process, and effectively realizes the editing and strengthening of the emotional gain to the sound signal;

3) During the processing of the sound signal, it is separated into human voice and background sound, and further gain processing is performed on the human voice, which is convenient for the hearing-impaired to recognize human voice without affecting their perception of environmental sound;

4) In the process of sound signal processing, two procedures of recognition and conversion are carried out for the human voice. Rereading can convert the human voice language with discordant components such as dialects into Mandarin which is easier to accept and recognize, which is convenient for understanding and reduces the The difficulty and threshold of learning;

5) The sound information output module fully covers sound signals of various spectrums through 48-channel filtering, so that the sound frequency range perceived by the hearing-impaired is consistent with that perceived by normal ears, avoiding additional discrimination or rejection.

Therefore, the present invention has a novel structure and ingenious conception. The emotional characteristics recognized by the computer are used to enhance the skin hearing effect, so that the hearing aid effect is better, the ability to receive sound frequencies is strong, and at the same time, it effectively reduces the difficulty and difficulty of learning for the hearing impaired. Threshold, the effect is remarkable.

Description of drawings

Fig. 1 is the overall structure schematic diagram of the present invention;

Fig. 2 is the schematic diagram of the sound preprocessing module of Fig. 1;

Fig. 3 is a schematic diagram of the information processing module of Fig. 1;

FIG. 4 is a schematic diagram of the sound information output module in FIG. 1 .

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings and specific preferred embodiments, but the protection scope of the present invention is not limited thereby.

A skin hearing aid device based on emotional gain, as shown in Figure 1-4, includes a microphone, a hearing aid host and electrode stickers, wherein the microphone can be set independently and connected to the hearing aid host through wires, or it can be set on the On the hearing-aid host, the electrode stickers use composite planar electrodes, including one or two in parallel, pasted on the skin surface behind the human ear, connected to the hearing-aid host through wires, and the hearing-aid host has a built-in power supply.

The hearing aid host includes a sound preprocessing module, an emotional analysis module, an emotional additional module, an information processing module and a sound information output module, wherein the sound preprocessing module is electrically connected to the microphone, including a digital module, a noise removal module and a sound correction module, and the digital module The analog signal output by the microphone is converted into a digital signal, the denoising module removes some inharmonic clutter to avoid affecting subsequent processing, the sound correction module adjusts the waveform; the emotion analysis module is connected to the sound preprocessing module and the emotion additional module, and the emotion analysis module It is connected to an emotional feature database; the information processing module is respectively connected to the sound preprocessing module, the emotional additional module and the sound information output module, and the information processing module processes the digital signal from the sound preprocessing module to make it convenient for disabled people to learn and train, and Send to the sound information output module, the sound information output module is respectively connected to the information processing module and the electrode sticker, the sound information output module converts the digital signal into an analog signal, and transmits it to the human skin through the electrode sticker; the sound information output module includes a filter and The booster, the information processing module, the filter, the booster and the electrode stickers are connected in sequence, the filter and the booster have several sets of parallel settings, the filter is a multi-channel bandpass filter, and each set of filters is selected separately Different center frequencies, the selection range is 15Hz~15kHz, 48 groups of filters and boosters are preferred, through different frequencies, 48 channels are used to output to the electrode stickers.

In the present invention, the processing of the digital signal from the sound preprocessing module by the information processing module includes successive steps of separation, recognition, addition, conversion, and merging. Signal separation, which separates the human voice from the background sound. The recognition link is to convert the human voice digital signal into an alphanumeric signal. The additional link is to receive the alphanumeric signal from the specific emotional characteristics sent by the emotional additional module. The conversion link is to convert the alphanumeric signal Perform rereading, convert it into a digital signal of simulated human voice, generally in Mandarin, combine specific emotional characteristics, adjust syllables and speech speed, and output a digital signal of simulated human voice with specific emotion. The sound digital signal is combined with the background sound digital signal, and the processed sound digital signal is output.

When the present invention is applied to the daily hearing aid of the disabled, after the microphone receives the sound, it converts it into an analog signal and inputs it to the hearing aid mainframe. Syllable correction, the sound digital signal is transmitted to the information processing module, at the same time, the emotion analysis module extracts the emotional features in the sound digital signal, compares it with the emotional feature database, determines it as the specified emotion, and transfers all the emotional features corresponding to the emotion to the emotion Additional module; after the information processing module obtains the sound digital signal, it separates the human voice from the background sound, and after adding the emotional features from the emotional additional module to the human voice, it rereads the digital signal of the simulated human voice, merges it with the background sound digital signal, and passes through The processed sound digital signal output is the sound information output module; the sound information output module is filtered by 48 channels of different frequency filters and amplified by the booster, and then output to the electrode sticker; the electrode stimulates the human skin, and the electric shock is transmitted to the human ear nerve , converted into human nerve impulse signals, transmitted through the nervous system, and projected to the auditory cortex of the brain, thereby generating hearing.

In the present invention, what people with disabilities hear is not the original sound, but the electronic sound processed by AI, which can easily remove the dissonant components in the human voice language, such as dialects, etc., and convert it into Mandarin, which is easier to accept and recognize. Therefore, for the hearing-impaired persons with congenital complete hearing loss, it can effectively reduce the difficulty and threshold of learning, but for the hearing-impaired persons with complete loss of hearing the day after tomorrow, since they have already learned the language and seen the original sound, they need to go through a relatively long process. short-term adaptation. But all in all, it can effectively solve the problem of hearing-impaired people who have completely lost their hearing, so that they can accurately hear the sounds from the world through hearing aids.

Therefore, in combination with the above structure and steps, it can be found that the skin hearing device based on emotion gain according to the present invention can gain the skin hearing effect through the emotional characteristics recognized by the computer, so that the hearing aid effect is better, and the receiving ability of the sound frequency is improved. At the same time, it effectively reduces the difficulty and threshold of learning for the hearing-impaired, and the effect is remarkable.

Claims (8)

1. A skin hearing aid based on emotional gain, comprising a microphone, electrode stickers, and a hearing aid host connected to the microphone and the electrode stickers, characterized in that: the hearing aid host includes a sound preprocessing module and an emotion analysis module , emotion additional module, information processing module and sound information output module, wherein the sound preprocessing module is electrically connected with the microphone, including a digitization module, the digitization module converts the analog signal output by the microphone into a digital signal, and the emotion analysis module is connected with the sound preprocessing module respectively It is connected with the emotional additional module, and the information processing module is respectively connected with the sound preprocessing module, the emotional additional module and the sound information output module, and the information processing module processes the digital signal from the sound preprocessing module to make it convenient for the disabled to learn and train, And sent to the sound information output module, the sound information output module is respectively connected to the information processing module and the electrode patch, and the sound information output module converts the digital signal into an analog signal, and transmits it to the human skin through the electrode patch; The processing of the digital signal from the sound preprocessing module by the information processing module includes successive steps of separation, identification, addition, conversion, and merging, wherein the separation step is to separate the input sound digital signal of the sound preprocessing module, The human voice is separated from the background sound. The recognition link is to convert the human voice digital signal into an alphanumeric signal. The additional link is to accept the alphanumeric signal to receive the emotional characteristics sent by the emotional additional module. The conversion link is to reread the alphanumeric signal and convert it For mandarin that simulates the human voice, adjust the syllable and speech rate in combination with the emotional features, and output the simulated human voice digital signal with emotion. Processed sound digital signal. 2. the skin hearing aid device based on emotional gain according to claim 1, characterized in that: the emotional analysis module is connected with an emotional feature database, extracts the emotional feature in the digital signal obtained by the digitization module, and the emotional feature The data is compared to determine the emotion contained in the language represented by the emotional feature expressed by the digital signal, and sent to the emotional additional module, and the corresponding emotional feature is superimposed on the digital signal processed by the information processing module through the emotional additional module. 3. The skin-acoustic hearing aid device based on emotional gain according to claim 1, characterized in that: the sound information output module includes a filter and a booster, wherein the information processing module, filter, booster The transformer and the electrode stickers are connected in sequence, and the filter and the booster are arranged in parallel in several groups. 4. the skin hearing aid device based on emotional gain according to claim 3, characterized in that: the filter is a multi-channel bandpass filter, and each group of filters selects different center frequencies respectively, and the selection range is 15Hz ~ 15kHz. 5. The skin-acoustic hearing aid device based on emotional gain according to claim 4, characterized in that: said filter and booster 48 groups, through different frequencies, use 48 channels to output to the electrode stickers. 6. The skin-acoustic hearing aid device based on emotional gain according to claim 1, characterized in that: the electrode stickers are connected to the hearing aid host through wires, and the electrode stickers are pasted on the skin where human nerves are densely distributed. 7. The skin-acoustic hearing aid device based on emotion gain according to claim 5, characterized in that: the electrode stickers are pasted on the surface of the skin behind the human ear. 8. The skin-acoustic hearing aid device based on emotion gain according to claim 1, characterized in that: the sound preprocessing module includes a denoising module and a syllable verification module.

Images