CN100589784C - Electrostimulation coding method for n degree marking method for Chinese language tone of electronic cochlea - Google Patents
Electrostimulation coding method for n degree marking method for Chinese language tone of electronic cochlea Download PDFInfo
- Publication number
- CN100589784C CN100589784C CN200610060582A CN200610060582A CN100589784C CN 100589784 C CN100589784 C CN 100589784C CN 200610060582 A CN200610060582 A CN 200610060582A CN 200610060582 A CN200610060582 A CN 200610060582A CN 100589784 C CN100589784 C CN 100589784C
- Authority
- CN
- China
- Prior art keywords
- electricity irritation
- classification
- fundamental frequency
- speed
- degree
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Abstract
An electric stimulation encoding method of electronic cochlea for the n-level marking method of Chinese tone includes such steps as measuring and fitting the resolution of deaf patient to electric stimulation rate, calculating a group of proper discrete electric stimulation rate values, classifying the received tone frequency information, adaptively choosing base frequency searching range, and searching the electric stimulation rate value associated with the classified tone frequency information to form a coding sequence of electric stimulation rate value.
Description
Technical field
The present invention relates to cochlear implant and recover deafness patient audition field, particularly a kind of adaptation feature of Chinese language is based on the electricity irritation coded method of the n scale notation of the cochlear implant speech processes of transmitting tone equifrequent information.
Background technology
Cochlear implant is also referred to as cochlear implant, electronic ear, bionic ear or artificial cochlea, and it is present unique device that can make anacusia patient recover audition.It relies on the physiological function of directly imitating peripheral auditory system with the excited acoustic fibers of weak current, produces the nerve similar to normal human ear and provides pattern, thereby recover patient's audition.
Cochlear implant can help anacusia patient to recover the communication ability, increases that they accept the education, the chance of employment and social communication.Especially no matter deaf youngster is that prelingual or language back are deaf, in the time still can't hearing sound by means of sonifer, will cause them to lose the chance that obtains normal education throughout one's life, increases white elephant for family and society.Cochlear implant can help them to regain tone sense, obtains extraneous information and knowledge, becomes the people useful to society.
The function of cochlear implant voice signal processing section be with acoustical signal be converted to cochlea in the corresponding electrical stimulation signal of electrode, the voice signal Processing Algorithm is the technology of most critical in the cochlear implant system, it has significant effects to the performance of cochlear implant.The algorithm that existing clinical prods adopts can be divided into two big classes, and a class is based on feature extraction, promptly extracts features such as the fundamental frequency of voice signal and formant, produces the stimulus signal of respective electrode then.One class is based on the voice signal Processing Algorithm of bank of filters, promptly voice signal is carried out the frequency-division section Filtering Processing.But the speech processing algorithm of existing clinical prods all is to develop on the basis that English is understood, and has mainly met Indo-European characteristics.And Chinese does not belong to Indo-European relatives' family of languages as one of Han-Tibetan family, and both differ greatly, and this also is to cause the existing voice Processing Algorithm be difficult to allow one of reason of China's patient satisfaction for the speech recognition effect of Chinese.
Chinese has the characteristic of oneself, and it is a kind of sound, rhyme, tone language, and wherein tone and intonation have great importance for the correct understanding of Chinese.Yet the product of clinical use only relies on the extraction of voice amplitude information and transmits the information such as tone that are difficult to effectively transmit Chinese, average recognition rate about 68% under its quiet condition, 99% ± 3.0% mutually far short of what is expected with the normal good hearing experimenter is difficult to satisfy patient's demand.
Therefore can make full use of the Chinese characteristic, the exploitation that more transmission Chinese language tones etc. enrich the cochlear implant speech processing algorithm of frequency information just seems especially important.The algorithm structure of speech processes can effectively utilize two parts according to what purposes was divided into the extraction of tone equifrequent information and this information, thus this wherein for Chinese frequency informations such as the tone that extracts how in electricity irritation in addition the efficient coding utilization effectively pass to the patient and just become the most important thing.
Summary of the invention
The objective of the invention is to the problems referred to above, a kind of electricity irritation coded method of n scale notation of Chinese language tone of electronic cochlea efficiently is provided, thereby recovers the audition of China's deafness patient more effectively at present cochlear implant existence.
The electricity irritation coded method of the n scale notation of Chinese language tone of electronic cochlea proposed by the invention is on the basis that the characteristics of electricity irritation audition are analyzed, and utilizes the result that frequency one tone is theoretical and method synthesis such as clinical quantized electricity irritation frequency discrimination capacity data and Chinese language tone five degree partitionings are studied.Coding strategy is divided into pretreatment module, preliminary classification module, the wide determination module of search degree and four parts of electricity irritation speed generation module, specifically may further comprise the steps:
A, pretreatment module are measured the also electricity irritation speed resolution capability of match deafness patient, calculate and can be differentiated by deafness patient and one group of discrete electricity irritation rate value of perception, with described one group of discrete electricity irritation rate value from slowly to being defined as 1 degree, 2 degree soon successively ... the n degree;
B, preliminary classification module are according to the tone equifrequent information classification of the preliminary classification standard that pre-establishes with the continuous variation of input reception;
C, sorted frequency information are by search width determination module, and adaptively selected fundamental frequency is searched for width;
D, electricity irritation speed generation module be according to selected fundamental frequency search width, constantly sorted tone equifrequent information and described some discrete electricity irritation rate values are carried out corresponding, thereby formation electricity irritation rate coding sequence.
The pretreatment of coding strategy mainly is measurement and match patient's an electricity irritation speed resolution capability.The electricity irritation reference speed rate adopts speed points such as [50 75 100 150 200 300 500 800 1000] Hz respectively during measurement, and is that the basis upwards changes electricity irritation speed with these reference speed rate, measures the resolution capability of the sentient tonal variations of patient.To carry out match to these measurement results after the measurement, be used for calculating the electricity irritation rate distribution that satisfies the distinguishable ability of patient, thereby lay the foundation for next step processing.According to clinical data in the past, patient's speed discrimination threshold increases along with the increase of electricity irritation speed, and both have exponential function relation.In order to guarantee that the patient can tell the rate variation of electricity irritation, must depict the upper limit of speed discrimination threshold with electricity irritation rate variation curve.Because both present exponential relationship, the present invention defines upper limit curve and is:
y＝A*e
B*x+R
A in the formula, B and R can adjust according to the clinical data that records.
Can calculate one group of discrete electricity irritation rate value being differentiated by the patient thus, the present invention's degree of being defined as value.In general, the degree value can be divided into about nine degree, and from spending to being defined as 1 degree, 2 soon respectively slowly ... 9 degree.
Enter the preliminary classification stage subsequently.Voice signal has generated level and smooth fundamental curve after extracting through fundamental frequency, carries out adaptive classification according to the size of initial fundamental frequency then.The principle of classification is formulated according to the fundamental frequency characteristics of different speaker basically, carried out the division of initial three classes respectively at the fundamental frequency scope of man, woman and child's voice, the initial electricity irritation speed that different classification is corresponding different, when carrying out the electricity irritation rate coding, just the part tone color of speaker has been carried out classification and transmitted like this.
After the preliminary classification, carried out the definition of the fundamental frequency search width of different classification according to the excursion of speaker fundamental frequency.The fundamental frequency excursion difference of the speaker pronunciation of classifying different, if use same search width to be difficult to adapt to the sound property of various different speaker, may reduce the dynamic range of certain classification, thereby influence coding effect, therefore, the present invention has designed the adaptive fundamental frequency search width definition that has realized at speaker tone color characteristics.
Fundamental frequency search width by definition, just tone frequency information such as the fundamental frequency of the variation that calculates and discrete electricity irritation rate value can be carried out correspondence, then by the speed of conversion electricity irritation along with the fundamental frequency variation, just the tone information with Chinese has passed to the patient, thereby helps the patient better to improve the Chinese identification ability.
The present invention uses for reference the relative pitch segment encoding method of Chinese language tone five scale notations, utilize match cochlear implant patient electricity irritation recognition threshold and tone sense to determine and to have been proposed the electricity irritation coding strategy of the Chinese language tone equifrequent information of innovation by the electricity irritation speed segmentation division methods of patient's perception.This coding strategy has realized that at speaker tone color characteristics classification is transmitted and the search of adaptive fundamental frequency, can keep and transmit the characteristic voice of speaker to greatest extent by limited electricity irritation degree value, be applicable to the coding transmission of Chinese frequency informations such as multiword mediation intonation and frequency modulation information, better improve the Chinese identification ability to help the patient.Its main effect is presented as:
A) in order better to transmit the tone color of speaker, improve speech discrimination score, the present invention has realized the part tone color of speaker has been carried out the algorithm that classification is transmitted by preliminary classification.Three preliminary classification have been represented the fundamental frequency scope of man, woman and child's voice respectively, keep and transmitted the characteristic voice of speaker to greatest extent by limited electricity irritation degree value.
B) the present invention has also carried out the definition of the fundamental frequency search width of different classification according to the excursion of speaker fundamental frequency, the every rising of fundamental frequency or reduce by one times search width, electricity irritation speed raises or once reduces, thus the adaptive fundamental frequency search that has realized at speaker tone color characteristics.
C) in order to guarantee that the patient can tell the rate variation of electricity irritation, must depict the upper limit of speed discrimination threshold with the rate variation curve.Because both present exponential relationship, the upper limit curve of exponential relationship that the present invention is self-defined, thus realized calculating discrete recognizable electricity irritation rate value.
D) because the multiword of Chinese is in harmonious proportion intonation also by this carrier perception of fundamental frequency, and the dynamic range of frequency modulation information and fundamental frequency are similar, so this coding strategy is equally applicable to multiword is in harmonious proportion the coding transmission of Chinese frequency informations such as intonation and frequency modulation information.
Description of drawings
Fig. 1 is the structured flowchart of this electricity irritation coding strategy;
Fig. 2 is the speed discrimination threshold upper limit and electricity irritation rate function fitted figure (abscissa is an electricity irritation speed among the figure, and vertical coordinate is the speed discrimination threshold, and curve 1 is an exponential curve of fit, and curve 2-9 is the matched curve of document measured data).
The specific embodiment
Below in conjunction with description of drawings the specific embodiment of the present invention.
Fig. 1 is an electricity irritation coding strategy structured flowchart of the present invention.It is divided into pretreatment module, preliminary classification module, the wide determination module of search degree and four parts of electricity irritation speed generation module.The specific coding strategy may further comprise the steps:
A, pretreatment module are measured the also electricity irritation speed resolution capability of match deafness patient, calculate and can be differentiated by deafness patient and one group of discrete electricity irritation rate value of perception, with described one group of discrete electricity irritation rate value from slowly to being defined as 1 degree, 2 degree soon successively ... the n degree;
B, preliminary classification module are according to the tone equifrequent information classification of the preliminary classification standard that pre-establishes with the continuous variation of input reception;
C, sorted frequency information are by search width determination module, and adaptively selected fundamental frequency is searched for width;
D, electricity irritation speed generation module be according to selected fundamental frequency search width, constantly sorted tone equifrequent information and described some discrete electricity irritation rate values are carried out corresponding, thereby formation electricity irritation rate coding sequence.
In step a, calculating can be by the method for one group of discrete electricity irritation rate value of deafness patient resolution and perception: definition patient speed discrimination threshold is with the upper limit curve y=A*e of electricity irritation rate variation
B*x+ R, in the formula, variable x is an electricity irritation speed, variable y is the speed discrimination threshold, according to document (
Zeng FG:Temporal pitch in electric hearing.Hearing Res.2002,174:101-106) the speed discrimination threshold of four cochlear implant patient 8 counter electrode of accurate measurement of record and the corresponding relation of electricity irritation speed calculate the value of parameter A, B, R; According to formula y=A*e
B*x+ R match upper limit curve on this upper limit curve, distinguishes that according to one group of speed determining threshold value obtains corresponding electricity irritation rate value, is can be differentiated by deafness patient and one group of discrete electricity irritation rate value of perception.
Present embodiment according to document (
Zeng FG:Temporal pitch in electric hearing.HearingRes.2002, 174:101-106) record eight measurement results, calculate upper limit curve y=A*e
B*xThe parameter of+R is A=8.3, B=0.0092, R=6.Its upper limit fitting result as shown in Figure 2, curve 1 is a upper limit curve among Fig. 2, curve 2-9 is the curve according to the corresponding relation match of the speed discrimination threshold of 8 counter electrode of above-mentioned document record and electricity irritation speed.The frequency discrimination threshold value that calculates according to this upper limit is used to refer to the interval of leading this paper setting stimulation rates.With the 50Hz stimulation rates is starting point, the frequency discrimination threshold value of calculating can get electricity irritation speed and be 50Hz the time according to upper limit curve 1 is 20Hz, therefore next electricity irritation speed that can implanted person's perceived pitch changes is 70Hz, calculating the frequency difference threshold of stimulation rates when being 70Hz according to upper limit curve again is 22Hz, next can be implanted the electricity irritation speed that changes of person's perceived pitch be 92Hz, the rest may be inferred, and present embodiment obtains that electronic cochlea implant persons can be differentiated and nine electricity irritation speed of the different interval of perception are:
[50?70?92?118?150?190?240?325?500]Hz
Like this, present embodiment just has been divided into electricity irritation speed nine speed of being differentiated by the patient of interval and perception, and present embodiment is defined as nine degree, and from spending to being defined as 1 degree, 2 soon respectively slowly ... 9 degree.
Enter the preliminary classification stage subsequently.Voice signal has generated level and smooth fundamental curve after extracting through fundamental frequency, carries out adaptive classification according to the size of initial fundamental frequency then.The principle of classification is formulated according to the fundamental frequency characteristics of different speaker, and present embodiment has carried out the division of initial three classes at the fundamental frequency scope of man, woman and child's voice, and specific classification standard is as follows:
Man's fundamental frequency scope has been represented in classification-1 basically, and woman's fundamental frequency scope has been represented in classification-2 basically, and the fundamental frequency scope of child's voice has been represented in classification-3 basically.The initial electricity irritation speed that different classification is corresponding different: classification-1 corresponding 4 degree; Classification-2 corresponding 5 degree; Classification-3 corresponding 6 degree.When carrying out the electricity irritation rate coding, just the part tone color of speaker has been carried out classification and transmitted like this.
After the preliminary classification, carried out the definition of the fundamental frequency search width of different classification according to the excursion of speaker fundamental frequency.Classification-1 and classification-3 are because the general variation of fundamental frequency little (about 50Hz), the search width is defined as 10Hz, i.e. the every rising 10Hz of fundamental frequency of classification-1 or classification-3, corresponding electricity irritation speed rising 1 degree, the every reduction of the fundamental frequency 10Hz of classification-1 or classification-3, corresponding electricity irritation speed reduces by 1 degree; Classification-2 is because fundamental frequency generally has the variable quantity about 100Hz, the search width is defined as 20Hz, promptly the fundamental frequency search width of classification-2 correspondences is 20Hz, the every rising 20Hz of fundamental frequency of classification-2, corresponding electricity irritation speed rising 1 degree, the every reduction of the fundamental frequency 20Hz of classification-2, corresponding electricity irritation speed reduces by 1 degree.Thereby the adaptive fundamental frequency search width definition that has realized at speaker tone color characteristics.
Fundamental frequency search width by definition, just tone frequency information such as the fundamental frequency of the variation that calculates and discrete electricity irritation rate value can be carried out correspondence, then by the speed of conversion electricity irritation along with the fundamental frequency variation, just the tone information with Chinese has passed to the patient, thereby helps the patient better to improve the Chinese identification ability.
Thus, the present invention uses for reference the relative pitch segment encoding method of Chinese language tone, utilizing match cochlear implant patient electricity irritation recognition threshold to determine can be by the electricity irritation speed segmentation division methods of patient's perception, the electricity irritation coding strategy of the Chinese language tone equifrequent information of innovation has been proposed, the perfect cochlear implant speech processing algorithm of suitable feature of Chinese language, generated the stimulating current that meets algorithm, thereby be that the audition of better recovering China patient is laid a good foundation.
Claims (4)
1, the electricity irritation coded method of the n scale notation of Chinese language tone of electronic cochlea is characterized in that may further comprise the steps:
A, pretreatment module are measured the also electricity irritation speed resolution capability of match deafness patient, calculate and can be differentiated by deafness patient and one group of discrete electricity irritation rate value of perception, with described one group of discrete electricity irritation rate value from slowly to being defined as 1 degree, 2 degree soon successively ... the n degree;
B, preliminary classification module are according to the tone frequency information classification of the preliminary classification standard that pre-establishes with the continuous variation of input reception;
C, sorted frequency information are by search width determination module, and adaptively selected fundamental frequency is searched for width;
D, electricity irritation speed generation module be according to selected fundamental frequency search width, constantly sorted tone frequency information and described some discrete electricity irritation rate values are carried out corresponding, thereby formation electricity irritation rate coding sequence.
2, the electricity irritation coded method of the n scale notation of Chinese language tone of electronic cochlea according to claim 1, it is characterized in that in step a calculating can be by the method for one group of discrete electricity irritation rate value of deafness patient resolution and perception: definition patient speed discrimination threshold is with the upper limit curve y=A*e of electricity irritation rate variation
B*x+ R, in the formula, variable x is an electricity irritation speed, variable y is the speed discrimination threshold, calculates the value of parameter A, B, R according to the corresponding relation of speed discrimination threshold and electricity irritation speed; According to formula y=A*e
B*x+ R match upper limit curve on this upper limit curve, distinguishes that according to one group of speed determining threshold value obtains corresponding electricity irritation rate value, is can be differentiated by deafness patient and one group of discrete electricity irritation rate value of perception.
3, the electricity irritation coded method of the n scale notation of Chinese language tone of electronic cochlea according to claim 1 is characterized in that described preliminary classification standard is to formulate according to the fundamental frequency characteristics of different speaker, and concrete standard is as follows:
Classification-1 is man's a fundamental frequency scope, and classification-2 is woman's a fundamental frequency scope, and classification-3 is the fundamental frequency scope of child's voice, and the initial electricity irritation speed of classification-1, classification-2 and classification-3 correspondences is respectively 4 degree, 5 degree and 6 degree.
4, the electricity irritation coded method of the n scale notation of Chinese language tone of electronic cochlea according to claim 3, the fundamental frequency search width that it is characterized in that described classification-1 and classification-3 correspondences is 10Hz, the every rising 10Hz of fundamental frequency of classification-1 or classification-3, corresponding electricity irritation speed rising 1 degree, the every reduction of the fundamental frequency 10Hz of classification-1 or classification-3, corresponding electricity irritation speed reduces by 1 degree; The fundamental frequency search width of classification-2 correspondences is 20Hz, the every rising 20Hz of fundamental frequency of classification-2, and corresponding electricity irritation speed rising 1 degree, the every reduction of the fundamental frequency 20Hz of classification-2, corresponding electricity irritation speed reduces by 1 and spends.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200610060582A CN100589784C (en) | 2006-05-15 | 2006-05-15 | Electrostimulation coding method for n degree marking method for Chinese language tone of electronic cochlea |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200610060582A CN100589784C (en) | 2006-05-15 | 2006-05-15 | Electrostimulation coding method for n degree marking method for Chinese language tone of electronic cochlea |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1861027A CN1861027A (en) | 2006-11-15 |
| CN100589784C true CN100589784C (en) | 2010-02-17 |
Family
ID=37388614
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200610060582A Expired - Fee Related CN100589784C (en) | 2006-05-15 | 2006-05-15 | Electrostimulation coding method for n degree marking method for Chinese language tone of electronic cochlea |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100589784C (en) |
-
2006
- 2006-05-15 CN CN200610060582A patent/CN100589784C/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN1861027A (en) | 2006-11-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1531722B (en) | Voice recognition system using implicit speaker adaptation | |
| Nie et al. | Encoding frequency modulation to improve cochlear implant performance in noise | |
| CN100488473C (en) | Full-implanting type artificial cochlea and method for making same | |
| Kates | Classification of background noises for hearing‐aid applications | |
| CN110072434A (en) | The sound acoustics biomarker used for assisting hearing equipment uses | |
| CN101716108B (en) | System for processing electronic cochlear signal and coding electric stimulation | |
| Lan et al. | A novel speech-processing strategy incorporating tonal information for cochlear implants | |
| CN105792745B (en) | Nerve action potential is detected using Sparse Signal Representation | |
| CN201532762U (en) | Simultaneous interpretation device special for individuals | |
| CN100563608C (en) | The electric cochlea Chinese fixed electric stimulation amplitude changing pattern in-vitro voice processing unit | |
| CN205900093U (en) | Make an uproar device and virtual reality equipment fall | |
| CN110368005A (en) | A kind of intelligent earphone and mood and physiological health monitoring method based on intelligent earphone | |
| CN109460144A (en) | A kind of brain-computer interface control system and method based on sounding neuropotential | |
| CN105765654A (en) | Hearing assistance device with fundamental frequency modification | |
| CN103892939B (en) | Improve language processing device for artificial cochlea and the method for Chinese tone recognition rate | |
| CN108711419A (en) | A kind of the ambient sound cognitive method and system of artificial cochlea | |
| CN104307100B (en) | A kind of method and system improving artificial cochlea's pitch perception | |
| KR20200122894A (en) | Apparatus and method for transmitting voice | |
| CN100589784C (en) | Electrostimulation coding method for n degree marking method for Chinese language tone of electronic cochlea | |
| CN109200469B (en) | Optimized coding method and system for enhancing tone perception of cochlear prosthesis | |
| Yost | Overview: psychoacoustics | |
| CN111150934B (en) | Evaluation system of Chinese tone coding strategy of cochlear implant | |
| CN205696110U (en) | A kind of sound transitions is sense of touch and Intelligent bracelet that applicable deaf mute wears | |
| EP4280628A1 (en) | Use of hearing instrument telecoils to determine contextual information, activities, or modified microphone signals | |
| CN111144482B (en) | Scene matching method and device for digital hearing aid and computer equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CP01 | Change in the name or title of a patent holder | ||
| CP01 | Change in the name or title of a patent holder |
Address after: Shenzhen Graduate School of Guangdong Province, Shenzhen City Xili 518055 Nanshan District University City Tsinghua University Patentee after: Shenzhen International Graduate School of Tsinghua University Address before: Shenzhen Graduate School of Guangdong Province, Shenzhen City Xili 518055 Nanshan District University City Tsinghua University Patentee before: GRADUATE SCHOOL AT SHENZHEN, TSINGHUA University |
|
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100217 |