WO2022080774A1 - Speech disorder assessment device, method, and program - Google Patents

Abstract

The present invention relates to a speech disorder assessment device, and comprises: a communication unit which receives recording data including an utterance voice recorded while performing at least one utterance task by a person subject to speech disorder assessment; and a control unit which assesses the received recording data on the basis of preconfigured criterion data, determines, on the basis of the result of the assessment, at least one of whether the person has a speech disorder, and the severity thereof, and outputs information indicating the result of the determination.

Description

Equine Disability Assessment Apparatus, Methods and Programs

The present invention relates to an apparatus for evaluating speech impairment, and more particularly, to an apparatus for determining whether a subject has speech impairment by evaluating data recorded and uttered by a subject.

Paralyzed speech disorder is a speech disorder in which muscle control of speech production mechanism is impaired due to damage to the central or peripheral nervous system, resulting in weakened or slowed muscle movement and poor coordination.

There is a problem in that in order to evaluate the presence and severity of a speech disorder such as paraplegic speech disorder, the person must directly speak face-to-face with an evaluator and receive an evaluation of the speech.

Accordingly, there is a need for a method in which the subject to be evaluated can be evaluated for the presence of speech disability and the severity of the disability without face-to-face with the evaluator, but the present invention for implementing this has not been disclosed.

The present invention for solving the above-described problems is to evaluate the recorded data of the subject's performance of the speech task based on standard data to determine whether the subject has speech impairment and the severity of the impairment.

In addition, the present invention utters and records a plurality of sentences with different meanings according to the reading position in the sentence, and evaluates the frequency, ratio, and length of pauses within or between words to determine whether the subject has speech impairment want to judge

In addition, the present invention builds learning data using the recorded data of the speaker's performance of the speech task, the evaluation result of the evaluator on the recorded data, and the speaker's information, and learns a judgment model based on this to determine whether speech is impaired. I want to create a model.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

Speech impairment evaluation apparatus according to an embodiment of the present invention for solving the above-described problems, a communication unit and a determination unit for receiving the recording data including the speech voice recorded by the speech disability evaluation subject performing at least one speech task and a control unit configured to evaluate at least one of whether the subject has speech impairment and the severity of speech impairment by evaluating the received recording data based on preset reference data.

In addition, the speech task is characterized in that it includes a sentence spacing reading task performed by recording the speaker's utterance of at least one sentence having a plurality of different meanings according to a spacing reading position in the sentence.

In addition, the control unit is characterized in that with respect to the recorded data of the sentence spacing reading task, at least one item of the frequency, ratio, and length of a pause within a word or between words is evaluated.

In addition, the control unit stores the recorded data and the determined result in a memory, and when the determined severity changes by more than a preset value from the previously determined severity by performing the utterance task, the control unit searches the memory for the severity It is characterized in that the extracted data corresponding to the cause of the numerical change is provided to the terminal of the subject.

In addition, the utterance task includes a maximum extended vocalization task, and the control unit transmits the recorded data of the subject including the spoken voice recorded by performing the maximum extended utterance task, the maximum extended utterance time, frequency change rate, amplitude change rate, caps It is characterized by analyzing and evaluating at least one item of Trump.

In addition, the speech task includes an alternating exercise task, and the control unit converts the recorded data of the subject including the speech voice recorded by performing the alternating exercise task to the speech speed, regularity of speech cycle, and regularity of speech amplitude. It is characterized in that it is evaluated by analyzing at least one of the items.

In addition, the speech task includes a paragraph reading task, and the control unit transmits the recorded data of the subject including the spoken voice recorded by performing the paragraph reading task to the speech speed, the frequency of the pause section, the ratio of the pause section, the pause It is characterized by analyzing and evaluating at least one of a section length, a fundamental frequency, and a cepstrum.

In addition, the control unit may receive an evaluation result of the evaluator on the recorded data including the speech voice recorded by the speaker performing the at least one speech task, and divide the recorded data of the speaker into a normal group based on the evaluation result Or classifying and storing it as a speech impaired group, and constructing learning data using at least one of the speaker's recorded data, the evaluation result of the speaker's recorded data, and the speaker's information to train the judgment model do.

In addition, when the speaker corresponds to a speech impairment, the control unit stores the cause of the speaker's speech disorder together, and further includes the cause of the speaker's speech disorder to build the learning data.

In addition, the control unit constructs big data using at least one of recorded data of speakers classified and stored as the normal group, evaluation results of each recorded data, and speaker information of each recorded data, and based on this, the standard data is characterized in that it is updated.

A speech impairment evaluation model construction system according to an embodiment of the present invention for solving the above-described problems provides at least one speech task for evaluating speech disability to a speaker terminal, and performs the at least one task to record A communication unit that receives the evaluation result of the evaluator for the recorded data of the speaker from the evaluator terminal, and classifies and stores the recorded data of the speaker into a normal group or a speech impaired group based on the evaluation result, the recorded data of the speaker, the and a control unit for constructing learning data using at least one of an evaluation result of the speaker's recorded data and information on the speaker, and learning the judgment unit based on the learning data.

Speech impairment evaluation method according to an embodiment of the present invention for solving the above-described problems, the steps of providing at least one speech task to the terminal of the subject for evaluation of speech disability, the subject performs the at least one speech task Receiving the recorded data including the recorded speech voice and evaluating the received recorded data through a determination unit based on preset reference data, and determining whether the subject has a speech impairment.

A method of constructing a speech impairment evaluation model according to an embodiment of the present invention for solving the above-described problems includes the steps of providing a speech task to a speaker terminal, receiving recording data of performing the speech task from the speaker terminal, and recording data It includes the steps of providing to the evaluator's terminal, receiving the evaluator's evaluation result, and classifying the recorded data into a normal group or a speech-impaired group based on the evaluation result and storing the same.

In addition to this, another method for implementing the present invention, another system, and a computer-readable recording medium for recording a computer program for executing the method may be further provided.

According to the present invention as described above, by evaluating the recorded data of the subject's performance of the speech task based on the standard data, there is an effect of determining whether the subject has speech impairment and the severity of the impairment.

In addition, according to the present invention, by uttering and recording a plurality of sentences having different meanings according to the spaced reading position in the sentence, and evaluating whether or not the subject has a speech impairment, the subject's speech using a Korean-specific sentence It has the effect of determining whether or not there is a disability.

In addition, according to the present invention, learning data is constructed using the recorded data of the speaker's speech task performance, the evaluation result of the evaluator on the recorded data, and the speaker's information, and based on this, a judgment model is trained to determine whether speech is impaired. It has the effect of being able to create a judging model.

Effects of the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the following description.

1 is a schematic diagram of a system for constructing a horse disability evaluation model according to an embodiment of the present invention.

2 is a schematic diagram of an apparatus for evaluating a horse's disability according to an embodiment of the present invention.

3 is a block diagram of an apparatus for evaluating a horse's disability according to an embodiment of the present invention.

4 to 16 are diagrams exemplifying that a speaker performs a speech task using a terminal and evaluates it by an evaluator.

17 is a flowchart of a method for constructing a horse disability evaluation model according to an embodiment of the present invention.

18 is a flowchart of a method for evaluating a horse's disability according to an embodiment of the present invention.

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only the present embodiments allow the disclosure of the present invention to be complete, and those of ordinary skill in the art to which the present invention pertains. It is provided to fully understand the scope of the present invention to those skilled in the art, and the present invention is only defined by the scope of the claims.

The terminology used herein is for the purpose of describing the embodiments and is not intended to limit the present invention. In this specification, the singular also includes the plural unless specifically stated otherwise in the phrase. As used herein, “comprises” and/or “comprising” does not exclude the presence or addition of one or more other components in addition to the stated components. Like reference numerals refer to like elements throughout, and "and/or" includes each and every combination of one or more of the recited elements. Although "first", "second", etc. are used to describe various elements, these elements are not limited by these terms, of course. These terms are only used to distinguish one component from another. Accordingly, it goes without saying that the first component mentioned below may be the second component within the spirit of the present invention.

Unless otherwise defined, all terms (including technical and scientific terms) used herein will have the meaning commonly understood by those of ordinary skill in the art to which this invention belongs. In addition, terms defined in a commonly used dictionary are not to be interpreted ideally or excessively unless specifically defined explicitly.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a schematic diagram of a equine disability evaluation model building system 10 according to an embodiment of the present invention, and FIG. 2 is a schematic diagram of a equine disability evaluation apparatus 100 according to an embodiment of the present invention.

With reference to FIGS. 1 and 2 , the system 10 for constructing a equine disability evaluation model according to an embodiment of the present invention will be described, and an apparatus for evaluating equine disability using the equine disability evaluation model constructed in this way will be described. to do it

The speech impairment evaluation apparatus 100 according to an embodiment of the present invention may be performed by an information processing device such as a computer, and specifically, since it is performed by communicating with the terminal 230 of the speaker, the evaluator, and the subject to be evaluated, speech evaluation It can be implemented as a server.

In addition, the speech evaluation server may mediate the speech evaluation of the speaker and the evaluator through a web or a service application, or may perform a speech evaluation of an evaluation target.

In the speech evaluation server, as shown in FIGS. 4 to 15 through the service application, the speaker performs a speech task through the terminal and records it, and the evaluator can evaluate the recorded data through the terminal, and a more detailed description will be made again below.

Referring to FIG. 1 , the speech impairment evaluation apparatus 100 sequentially provides at least one speech task to the talker terminal 210 , wherein the speech task is provided in the form of text or an image to the screen of the talker terminal 210 . It is preferable to output as

The speaker performs the presented speech task to record the speech voice, and the speech impairment evaluation apparatus 100 receives the recorded data recorded by performing the speech task from the talker terminal 210 .

The evaluator accesses and connects to the speech disability evaluation device 100 through the terminal, receives or streaming playback of the speaker's recorded data, listens to the recorded data and performs evaluation to input the evaluation result into the terminal, and the speech disability evaluation device ( 100 receives the evaluation result from the evaluator terminal 220 .

The speech disability evaluation apparatus 100 classifies the information of the speaker, the recorded data of the speaker, and the evaluation result of the recorded data into a normal group and a speech impairment group according to the evaluation result of the evaluator, and stores the classified information in the memory 130 . Then, learning data is constructed by using at least one of the speaker's recorded data, the evaluation result of the speaker's recorded data, and the speaker's information, and a judgment model is trained based on this.

In addition, the speech disability evaluation apparatus 100 uses at least one of the speaker information, the speaker's recorded data, and the evaluation result of the recorded data classified as a normal group to update the standard data stored in the memory 130 . can

In some embodiments, the speech disability evaluation apparatus 100 may generate standard data for evaluating speech disability by using at least one of information of a speaker classified as a normal group and stored, recorded data of the speaker, and an evaluation result of the recorded data. may be

These standard data may be provided to and utilized by the evaluator terminal 220 , or may be stored in the memory 130 and become a criterion for the control unit 110 to evaluate the subject's speech impairment.

The speech disability evaluation apparatus 100 performs the above process for a plurality of speakers and builds big data using the data obtained, and by building the learning data to learn the judgment model, a model capable of self-evaluating speech disability will build

Referring to FIG. 2 , the speech impairment evaluation apparatus 100 provides at least one speech task to the terminal 230 of the subject for speech impairment evaluation, and the subject performs the speech task to record the speech performance voice.

In the embodiment described below, the speech impairment evaluation apparatus 100 performs a speech task with the terminal 230 of the speech impairment evaluation target to record the speech voice, and the control unit evaluates the recorded data and outputs the evaluation result. Although described as that, it is not limited to the terminal of the subject.

For example, the subject of speech disability evaluation may use the speech disability evaluation apparatus 100 to perform speech disability evaluation.

To this end, the speech impairment evaluation apparatus 100 outputs a speech task through the output unit 150 (eg, a display module, a voice output device, etc.), and a voice input device (eg, a microphone) that allows the subject to input a voice. ) may be further included as a component.

The speech impairment evaluation apparatus 100 receives the recording data from the terminal 230 of the subject through the communication unit 120 , and the control unit 110 uses the judgment model stored in the memory 130 to set the recorded data as standard data. based on the evaluation, and based on the evaluated result, at least one of whether or not the subject has speech impairment and the severity of the speech impairment is evaluated.

More specifically, each speech task includes a plurality of evaluation items.

Then, the controller 110 evaluates the spoken voice recorded in the recorded data according to the utterance task, and evaluates a plurality of evaluation items of each utterance task based on preset reference data.

In an embodiment of the present invention, the speaker, the evaluator, and the terminal of the subject are a microphone for inputting a voice, a display unit for outputting a speech task, a speaker capable of outputting recorded data, the speech impairment evaluation apparatus 100 and data Any device having the communication unit 120 capable of performing communication is applicable, and representative devices such as a smartphone, a tablet PC, and a notebook computer are applicable.

Hereinafter, it will be described in more detail with reference to other drawings.

3 is a block diagram of the apparatus 100 for evaluating a horse's disability according to an embodiment of the present invention.

4 to 15 are diagrams exemplifying that a speaker performs a speech task using a terminal and evaluates it by an evaluator.

First, with reference to Figure 3 will be described with respect to the horse disability evaluation model building system 10 according to an embodiment of the present invention.

Referring to FIG. 3 , the speech impairment evaluation apparatus 100 includes a control unit 110 , a communication unit 120 , a memory 130 , a voice recognition module 140 , and an output unit 150 .

However, in some embodiments, the horse disability evaluation apparatus 100 may include a smaller number of components or more components than the components shown in FIG. 3 .

The communication unit 120 provides at least one speech task for the evaluation of speech impairment to the speaker terminal 210, and provides an evaluation result of the evaluator for the recorded data of the speaker recorded by performing the speech task to the speaker terminal 210 do.

The memory 130 of the speech impairment evaluation apparatus 100 may store speech tasks for evaluating whether the speaker has speech impairment.

In an embodiment of the present invention, the memory 130 may be provided as a component in the horse disability evaluation apparatus 100, or may be provided in the same form as an external cloud server.

The control unit 110 classifies the speaker's recording data into a normal group or a speech impairment group based on the evaluation result received from the evaluator terminal 220 and stores it in the memory 130 .

In more detail, the controller 110 classifies at least one of the speaker information, the speaker's recorded data, and the evaluation result of the recorded data into a normal group or a speech impairment group based on the evaluation result of the evaluator on the recorded data. and stored in the memory 130 .

Also, the controller 110 may store the cause of the speaker's speech disorder and the severity of speech disorder together based on the evaluation result.

Then, the controller 110 builds learning data by using at least one of the speaker's information, the speaker's recorded data, and the evaluation result of the recorded data, and learns the judgment model stored in the memory 130 based thereon.

In this case, when the speaker corresponds to a speech disability based on the evaluation result, the controller 110 may further include the cause of the speaker's speech disorder and the severity of the speech disorder in the learning data.

With this configuration, the speech disability evaluation device 10 builds learning data by including the speaker's speech disorder cause and speech disorder severity together, and by learning the judgment model, the subject based on the recording data of the subject for subsequent speech impairment evaluation It is possible to evaluate not only whether the person has a speech impairment, but also the cause and severity of the speech impairment.

In one embodiment, the speech impairment evaluation apparatus 100 includes a speech recognition module 140 having a speech recognition function, and the controller 110 controls the speech recognition module 140 to record data of performing a speech task. By recognizing the voice, the judgment unit can perform the analysis, and the control unit 110 builds learning data and inputs it to the voice recognition module 140 to learn it, so that the voice recognition accuracy of a person with a speech impairment can be improved. .

In this way, the voice recognition module 140 can recognize a voice from the recorded data, and recognize the voice of a person with a speech impairment as a result of learning through the recorded data performed by a plurality of speakers, evaluation results, and information of the speakers. Accuracy is guaranteed.

In one embodiment, the controller 110 constructs big data using at least one of recorded data of speakers classified and stored as a normal group, evaluation results for each recorded data, and speaker information of each recorded data, and based on the to update the normative data.

In this case, the standard data refers to a criterion for evaluating whether speech is impaired for a plurality of evaluation items with respect to the recorded data, and an evaluation criterion for evaluating a score.

Accordingly, this means that the standard data previously used is stored in the memory 130 , and the control unit 110 updates various reference values for a plurality of evaluation items of the norm data using big data.

4 to 15 , it will be described that the speech disability evaluation apparatus 100 according to an embodiment of the present invention performs non-face-to-face speech disability evaluation by intermediarying the talker terminal 210 and the evaluator terminal 220 . .

As mentioned above, the equine disability evaluation apparatus 100 provides a equine disability evaluation method through a service application, and the speaker, the evaluation target, and the evaluator each install and sign up the service application in their own terminal, log in, and provide the service. is available.

When the evaluator accesses the evaluator's account through the terminal and activates the evaluation mode or the evaluation page, the evaluator may select at least one from the list of speakers waiting for evaluation and perform evaluation.

In this case, the evaluator's evaluation of the speaker's speech task performance recorded data may be performed in real time by connecting the evaluator and the speaker at the same time, or may be performed on recorded data previously performed and stored in the memory 130 .

4 exemplifies that the speaker and the subject to be evaluated install the service application in the terminal and input various information in the process of subscribing.

In an embodiment of the present invention, as essential information for evaluating whether the speaker or subject has speech impairment, at least one of name, gender, language, date of birth, academic background, and number of years of education may be applied. can be reflected.

And, in an embodiment of the present invention, as the selection information for evaluating whether the speaker and the subject have speech impairment, registration number, handle (Handedness), hearing, vision, teeth, medical diagnosis, communication disorder diagnosis name, elapsed day after the onset , main complaint, medications being taken, mental status information, etc. can be applied.

Accordingly, when the controller 110 constructs the learning data, the learning data may be constructed by including the above information together, and the information included in the learning data may be determined by the controller 110 according to each situation.

The speech task used in the embodiment of the present invention is the maximum extended vocalization task, the alternating movement task, the paragraph reading task, and the sentence spacing task, and will be described in order with reference to the drawings.

5 exemplifies that a maximum extended phonation task is given to the terminal, and the maximum extended phonation task is to measure a maximum phonation time (MPT), specifically 'Ah' prolongation (a long utterance of 'ah') below) is applicable.

The speaker activates the recording function by inputting the record button on the terminal screen, and performs the speech task by uttering 'a' for as long as possible.

The measurement of the maximum extended vocalization time can be repeated a plurality of times, and the maximum performance is reflected in the calculation of the result and standard score.

6 is an example of the evaluator evaluating the recorded data of the speaker's performance of the maximum extended vocalization task, and scoring items such as pitch, loudness, and quality are exemplified, and other than the evaluation items It is exemplified that other opinions of

Diadochokinetic rate (DDK) tasks include alternating motion rate (AMR) tasks that record utterances that are continuously repeated for a specific monosyllable, and utterances that the speaker continuously repeats for two or more different syllables. to include at least one of the Sequential Motion Rates (SMR) tasks to record

In an embodiment of the present invention, the AMR task may perform the utterance task by continuously repeating specific monosyllables such as putter, turkey, and ker.

7 exemplifies that the task of measuring the alternating motion speed (AMR) is given to the terminal, and the talker activates the recording function by inputting the record button, and repeatedly utters 'fur' to perform the utterance task.

In addition, the horse disability evaluation apparatus 100 may request the measurement of the alternating movement speed for the tucker and the kerchief.

FIG. 8 exemplifies that the task of measuring serial motion speed (SMR) is given to the terminal, and the speaker inputs the record button to activate the recording function, and repeatedly utters 'putucker' to perform the speech task.

In one embodiment, the alternating motion task is performed for a preset time (eg, 5 seconds), and the speed of the lowest speed (Rate) and the sequential motion speed (SMR) during the measurement of the alternating motion speed (AMR) is calculated as a result and a standard score is reflected in

9 is an example of an evaluator evaluating recording data of a speaker's performance of an alternating motor task. The evaluator listens to the speaker's recorded data, and whenever the speaker repeats a specific speech sound, presses the measurement button of the terminal, the number on the screen is counted and increased, and the speed can be measured through this.

FIG. 10 exemplifies that the horse disability evaluation apparatus 100 provides the speaker terminal 210 to utter 'beach' in the serial motion speed (SMR) measurement task.

Speech impairment was evaluated using 'Buttercup' in an alternating exercise task for subjects using the English language, but there was a problem in that there was no corresponding suggested word in Korean.

The horse disability evaluation apparatus 100 according to an embodiment of the present invention provides a suggestion word 'beach' in a series of motion speed measurement task, and makes the speaker continuously repeat utterance 'beach' for a preset period of time and records, and the evaluator will be evaluated through this.

When SMR is performed through the suggestion word 'beach', the change of articulation position in the oral cavity (anterior → middle → posterior) appears during one cycle like 'putucker', but unlike 'putucker', it is a meaningful word, It has the advantage of not requiring parasitism (quality due to the friction of the glottis) at the time of calculation.

Therefore, if SMR is performed through the suggestion word 'beach', the effect of performing the task is excellent for patients who do not understand the utterance task using 'putucker' (eg, preschool children) in that they use words with meaning. There is an advantage that

In addition, the fact that SMR is not required at the time of calculation by performing SMR through the suggestion word 'beach' is that the consonants of parasitic qualities (for example, 'P', 'T', 'ㅋ') due to the deterioration of vocal cord function. ), it has the advantage of being easy for patients who cannot perform the task using the suggested word 'putucker'.

The evaluator listens to the recorded data on the performance of the alternating exercise task, and can evaluate evaluation items such as regularity, sound quality, and resonance, and opinions other than the evaluation items can be entered in a memo.

More specifically, in an embodiment of the present invention, the evaluator may auditory perceptually evaluate the regularity of the speaker by using an interval scale of a preset score for the regularity evaluation item.

In addition, the evaluator can auditory perceptually evaluate the breathy, strained, and continuous or intermittent of the speaker by using the interval scale of preset scores for the sound quality evaluation items.

In addition, the evaluator can auditory-perceptually evaluate the speaker's hyponasality, hypernasality, and continuous or intermittent with respect to the resonance evaluation item using an interval scale of a preset score.

11 exemplifies that a paragraph reading measurement task is given to the terminal, the speaker inputs the record button to activate the recording function, and reads the given sentence to perform the speech task.

In this case, as the paragraph reading task, a short paragraph reading task of less than a preset length or a long paragraph reading task of a preset length or more may be given, and a paragraph of a corresponding language is provided according to the language selection.

As shown in Figure 12, the evaluator can measure the speed for each section through the recorded data on the paragraph reading task performance and measure the reading speed to compare the speed change through this, and speech (utterance) intelligibility, resonance, and breathing problems Evaluation items such as evaluation items can be evaluated, and opinions other than evaluation items can be entered in a memo.

13 exemplifies that a sentence reading task is given by a terminal, the speaker inputs a record button to activate a recording function, and utters a given sentence to perform the utterance task.

More specifically, in the case of Korean, even in a sentence composed of the same syllables, there are a plurality of sentences having different meanings depending on the spacing/reading position in the sentence.

Therefore, the speech impairment evaluation apparatus 100 according to an embodiment of the present invention performs a sentence spacing reading task performed by recording the speaker's utterance for at least one sentence having a plurality of different meanings according to the spacing reading position in the sentence. In this way, the speaker and the subject's speech impairment will be evaluated.

The present applicant studied the Korean language-specific test method from the existing speech impairment evaluation method, and when the above sentences are read and evaluated, the result was obtained that it was possible to effectively evaluate the presence and severity of speech impairments of the speaker and the subject.

In Korean, 'eojeol' is each word that composes a sentence, and it is a unit of spacing/reading. It can be difficult to understand the meaning.

In particular, in the case of a language that must be pronounced with a clear gap between words, such as Korean, this difficulty is conspicuous.

As an embodiment, the speech impairment evaluation apparatus 100 may provide a sentence as shown in FIG. 13 as a sentence spacing reading task.

Referring to FIG. 13 , the first sentence (“lose me only once”) and the second sentence (“touch me once”) have completely different meanings depending on the positions of writing/reading spaces.

The speaker enters the record button to activate recording, reads the first sentence to record the spoken voice, and then completes it, then inputs the record button again to activate recording, reads the second sentence to record the spoken voice, Recorded data of each of the first sentence and the second sentence may be generated.

In this case, as shown in FIG. 13 , the evaluator may evaluate how the evaluator reads two sentences by presenting different first and second sentences having a plurality of meanings, or presenting only one sentence to perform the speech task. .

The evaluator may evaluate at least one item among the frequency, ratio, and length of pauses within or between words with respect to the recorded data of performing the sentence spacing reading task. In addition, an evaluation item corresponding to whether the speaker has spoken so that the correct meaning of the first sentence and the second sentence is transmitted may be added.

14 to 16 exemplify the output of the evaluation result of the evaluator with respect to the recorded data of the speaker's performance of the speech task.

Through the above processes, the speaker can perform a speech impairment evaluation non-face-to-face with the evaluator through the terminal, and the evaluation result is also provided to the terminal and confirmed.

And, as described above, the speech impairment evaluation apparatus 100 mediates the speech impairment evaluation of the speaker and the evaluator in this way, and learns the judgment model using numerous result data.

Hereinafter, as described above, the horse disability evaluation apparatus 100 uses the judgment model constructed by the horse disability evaluation model building system 10 to describe the evaluation of the horse disability evaluation target.

The speech impairment evaluation apparatus 100 provides a speech task to the terminal 230 of the subject to be evaluated, and receiving the recorded data may be the same as targeting the speaker described above, and the role of the evaluator is The evaluation device 100 takes over.

Subjects described below refer to subjects to be evaluated for performing speech impairment evaluation.

The control unit 110 of the speech impairment evaluation apparatus 100 provides at least one speech task to the terminal 230 of the subject through the communication unit 120 and requests the subject to record the speech.

The communication unit 120 receives the recorded data including the speech voice recorded by the subject performing at least one speech task.

The control unit 110 evaluates the received recording data based on preset reference data using the judgment model, and determines at least one of whether or not the subject has speech impairment and the severity of the speech impairment based on the evaluation result.

Then, the control unit 110 outputs information indicating the determined result.

In this case, the control unit 110 may output information through the output unit 150 (eg, a display module, a voice output device, etc.) provided in the speech impairment evaluation apparatus 100 , or information through the terminal 230 of the subject. can also be printed out.

In this case, as described above, the speech task may include at least one speech task among a maximum extended vocalization task, an alternating exercise task, a paragraph reading task, and a sentence spacing task.

Embodiments described below relate to a plurality of evaluation items included in each of the above-mentioned speech tasks. The control unit 110 evaluates the recorded data of performing each speech task included in the recorded data for the following evaluation items, and the standard data is applied as the evaluation criteria at this time.

In one embodiment, the control unit 110, with respect to the recorded data including the spoken voice recorded by the subject performing a sentence spacing reading task, at least one item of the frequency, ratio, and length of the inter-word or inter-word rest period. It is characterized in that it is analyzed and evaluated.

In one embodiment, the controller 110 analyzes the recorded data including the spoken voice recorded by the subject performing the maximum extended vocalization task with respect to at least one item of the maximum extended vocalization time, the frequency change rate, the amplitude change rate, and the cepstrum. It is characterized by evaluation.

In one embodiment, the control unit 110 analyzes recorded data including speech voice recorded by the subject performing an alternating exercise task for at least one item among speech speed, regularity of speech cycle, and regularity of speech amplitude. It is characterized by evaluation.

In one embodiment, the control unit 110 transmits the recorded data including the speech voice recorded by the subject performing the paragraph reading task to the speech speed, the frequency of the pause section, the ratio of the pause section, the length of the pause section, the fundamental frequency, and the caps. It is characterized by analyzing and evaluating at least one item of Trump.

The control unit 110 of the speech impairment evaluation apparatus 100 evaluates at least one speech task performed by the subject as above, and provides or outputs the evaluation result to the subject's terminal 230 through the communication unit 120 .

The control unit 110 may store the recording data and the determination result of the subject in the memory 130 and provide a management service for the subject's speech impairment and severity.

In one embodiment, the controller 110 is configured to control the subject's speech impairment determined as a result of the speech impairment evaluation conducted this time when the severity of the subject's speech impairment has changed by more than a preset value compared to the severity determined by performing the previous speech task, (if worsened, improved may include all cases) may be retrieved from the memory 130 to extract the recording data corresponding to the cause of the change in severity level and provided to the terminal 230 of the subject.

In the case of simply evaluating and reporting whether or not the subject's speech impairment and severity level are present, there is no way for the subject and the subject's guardian to know why the subject's speech impairment has improved or has worsened.

However, the speech impairment evaluation apparatus 100 according to the embodiment of the present invention provides the above-described configuration, by searching and providing recorded data of the existing speech performance matching the change in severity due to this evaluation, the subject or the subject's It has the effect of allowing the guardian to directly know the reason for the change in the severity level by listening and comparing the two recorded data (the previous recording data and the current recording data).

In one embodiment, the control unit 110 provides an evaluation result for the existing recording data and an evaluation result for the current recording data together, provides a comparison result for a plurality of evaluation items, or separately intermediaries an evaluator and performs two recordings A function of providing data to the evaluator terminal 220 to receive a detailed diagnosis may be provided.

Speech recognition technology currently used is focused on recognizing the voice of a normal person, and the accuracy of recognizing the speech or utterance of a patient with speech impairment (eg, paraplegic speech disorder) with poor speech intelligibility (accuracy) is remarkably high. There is a problem with falling.

In addition, in order for a patient with a speech impairment (especially paraplegic speech impairment) to evaluate and examine the presence and severity of a speech impairment, a subject who has or is presumed to have a speech impairment performed the evaluation face-to-face with the evaluator. There were difficulties in that there were subjects who were reluctant to face inconvenience and face-to-face.

Although these difficulties exist, only the speech recognition technology of the general public (normal group) has been developed, and the development of the technology for recognizing the speech of people with speech disabilities is remarkably lagging behind.

However, as described above, the speech impairment evaluation apparatus 100 according to an embodiment of the present invention uses the speech performance recording data of a plurality of speakers and the evaluation result of the evaluator to obtain an analysis model capable of evaluating speech impairment and severity. It has the effect that it can be applied to subjects who want to actually evaluate speech impairment using the built-up analysis model.

17 is a flowchart of a method for constructing a horse disability evaluation model according to an embodiment of the present invention.

Referring to FIG. 17, a method for constructing a horse disability evaluation model according to an embodiment of the present invention will be described.

First, the control unit 110 of the speech impairment evaluation apparatus 100 provides a speech task to the terminal 210 of the speaker through the communication unit 120 . (S110)

S100 Next, the control unit 110 receives the recording data of performing the speech task through the communication unit 120 . (S120)

S120 Next, the control unit 110 provides the recording data received in S120 through the communication unit 120 to the evaluator's terminal 220 . (S130)

S130 Next, the control unit 110 receives the evaluation result of the evaluator through the communication unit 120 . (S140)

S140 Next, the control unit 110 classifies and stores the speaker's task performance recording data into a normal group or a speech impaired group based on the evaluation result. (S150)

The equine disability evaluation model construction method according to the embodiment of the present invention described above differs from the equine disability evaluation model construction apparatus described with reference to FIGS. 1 to 16 only in the category of the invention, and since the content is the same, overlapping descriptions and examples will be omitted. do.

18 is a flowchart of a method for evaluating a horse's disability according to an embodiment of the present invention.

Referring to FIG. 18, a method for evaluating a horse's disability according to an embodiment of the present invention will be described.

First, the control unit 110 of the apparatus 100 for evaluating speech disability provides at least one speech task to the terminal 230 of the subject to be evaluated through the communication unit 120 . (S310)

S310 Next, the communication unit 120 receives the recording data of the speech task performance from the terminal 230 of the subject. (S330)

S330 Next, the control unit 110 evaluates the recording data received in S330 based on preset reference data, and determines at least one of whether or not the subject has speech impairment and the severity of the speech impairment based on the evaluation result. (S350)

The equine disability evaluation method according to the embodiment of the present invention described above differs from the equine disability evaluation apparatus 100 described with reference to FIGS. 1 to 16 only in the category of the invention, and since the content is the same, overlapping descriptions and examples will be omitted. .

The method according to an embodiment of the present invention described above may be implemented as a program (or application) to be executed in combination with a server, which is hardware, and stored in a medium.

The above-described program is C, C++, JAVA, machine language, etc. that a processor (CPU) of the computer can read through a device interface of the computer in order for the computer to read the program and execute the methods implemented as a program It may include code (Code) coded in the computer language of Such code may include functional code related to a function defining functions necessary for executing the methods, etc., and includes an execution procedure related control code necessary for the processor of the computer to execute the functions according to a predetermined procedure. can do. In addition, this code may further include additional information necessary for the processor of the computer to execute the functions or code related to memory reference for which location (address address) in the internal or external memory of the computer should be referenced. there is. In addition, when the processor of the computer needs to communicate with any other computer or server located remotely in order to execute the functions, the code uses the communication module of the computer to determine how to communicate with any other computer or server remotely. It may further include a communication-related code for whether to communicate and what information or media to transmit and receive during communication.

The storage medium is not a medium that stores data for a short moment, such as a register, a cache, a memory, etc., but a medium that stores data semi-permanently and can be read by a device. Specifically, examples of the storage medium include, but are not limited to, ROM, RAM, CD-ROM, magnetic tape, floppy disk, and optical data storage device. That is, the program may be stored in various recording media on various servers accessible by the computer or in various recording media on the computer of the user. In addition, the medium may be distributed in a computer system connected to a network, and a computer-readable code may be stored in a distributed manner.

The steps of a method or algorithm described in relation to an embodiment of the present invention may be implemented directly in hardware, as a software module executed by hardware, or by a combination thereof. A software module may contain random access memory (RAM), read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, hard disk, removable disk, CD-ROM, or It may reside in any type of computer-readable recording medium well known in the art to which the present invention pertains.

As mentioned above, although embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art to which the present invention pertains know that the present invention may be embodied in other specific forms without changing the technical spirit or essential features thereof. you will be able to understand Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive.

[Explanation of code]

10: Equine disability evaluation model building system

100: speech disability evaluation device

110: control unit

120: communication department

130: memory

140: voice recognition module

150: output unit

210: talker terminal

220: evaluator terminal

230: target terminal

Claims (13)

1. a communication unit configured to receive recorded data including speech voice recorded by the speech impairment evaluation subject performing at least one speech task; and

   A control unit that evaluates the received recording data based on preset reference data, determines at least one of whether or not the subject has speech impairment and the severity of the subject based on the evaluated result, and outputs information indicating the determined result containing,

   Equine Disability Assessment Device.
2. According to claim 1,

   The speech task comprises a sentence spacing reading task, performed by recording the speaker's utterance with respect to at least one sentence having a plurality of different meanings according to a spacing reading position in the sentence,

   Equine Disability Assessment Device.
3. 3. The method of claim 2,

   Wherein the control unit evaluates at least one item among the frequency, ratio, and length of a pause within a word or between words with respect to the recorded data of the sentence spacing reading task,

   Equine Disability Assessment Device.
4. According to claim 1,

   The control unit stores the recorded data and the determined result in a memory, and when the determined severity changes by more than a preset value from the previously determined severity by performing the utterance task, the control unit searches the memory to change the severity value Characterized in extracting the recording data corresponding to the cause of the subject and providing it to the terminal of the subject,

   Equine Disability Assessment Device.
5. According to claim 1,

   The speech task includes a maximum extended vocal task,

   The control unit analyzes and evaluates the recorded data of the subject including the spoken voice recorded by performing the maximum extended vocalization task with respect to at least one of a maximum extended vocalization time, a frequency change rate, an amplitude change rate, and a cepstrum to do,

   Equine Disability Assessment Device.
6. According to claim 1,

   The speech task includes an alternating motor task,

   The control unit analyzes and evaluates the recorded data of the subject including the speech voice recorded by performing the alternating exercise task for at least one of speech speed, regularity of speech cycle, and regularity of speech amplitude to do,

   Equine Disability Assessment Device.
7. According to claim 1,

   The speech task includes a paragraph reading task,

   The control unit transmits the recorded data of the subject, including the speech voice recorded by performing the paragraph reading task, at least one of a speech speed, a frequency of a pause section, a ratio of a pause section, a length of a pause section, a fundamental frequency, and a cepstrum. Characterized in analyzing and evaluating the items,

   Equine Disability Assessment Device.
8. According to claim 1,

   The control unit is

   The speaker performs the at least one speech task to receive an evaluation result of the evaluator on the recorded data including the recorded speech voice, and classifies the speaker's recorded data into a normal group or a speech impaired group based on the evaluation result and save it,

   characterized in that the judgment model is learned by constructing learning data using at least one of the speaker's recorded data, the evaluation result of the speaker's recorded data, and the speaker's information,

   Equine Disability Assessment Device.
9. 9. The method of claim 8,

   When the speaker corresponds to a speech impairment, the control unit stores the cause of the speaker's speech disorder together, and further includes the cause of the speaker's speech disorder to build the learning data,

   Equine Disability Assessment Device.
10. 9. The method of claim 8,

    The control unit builds big data using at least one of recorded data of speakers classified and stored as the normal group, evaluation results of each recorded data, and speaker information of each recorded data, and updates the standard data based on this. characterized in that

    Equine Disability Assessment Device.
11. a communication unit that provides at least one speech task for evaluation of speech impairment to a speaker terminal, and receives an evaluation result of the evaluator for recording data of the speaker recorded by performing the at least one task from the evaluator terminal; and

    Classifying and storing the recorded data of the speaker as a normal group or a speech impaired group based on the evaluation result, and using at least one of the speaker's recorded data, the evaluation result of the speaker's recorded data, and the speaker's information Containing a control unit for building the learning data and learning the judgment model based on the learning data,

    Equine disability assessment model building system.
12. performed by the Equine Disability Assessment Apparatus;

    providing at least one speech task to a terminal of a subject for evaluation of speech impairment;

    receiving recorded data including spoken voice recorded by the subject performing the at least one speech task;

    determining at least one of whether or not the subject has speech impairment and severity by evaluating the received recording data based on preset reference data; and

    Including the step of outputting information indicating the determined result,

    Methods for evaluating equine disability.
13. A recording medium having a program for executing the method of claim 12 in combination with a computer as hardware.

Images