JPH05119792A - Speech recognition device - Google Patents

Abstract

PURPOSE:To provide the speech recognition device which is improved in speech recognition performance at emergency time without spoiling speech recognition performance at normal time. CONSTITUTION:This speech recognition device is provided with an emergency detecting means 1 which detects emergency from an input speech, an input speech pattern generating means 2 which generates an input speech pattern by analyzing the input speech, a registered speech standard pattern storage means 5 which is stored with plural previously registered speech standard patterns, a matching part 3 which finds the distance of the input speech pattern to each registered speech standard pattern by matching the input speech pattern with the registered speech standard patterns, and a recognition discrimination means 4 which discriminates the recognition result according to the pattern-to- pattern distance; and the speech recognizing process is performed corresponding to the emergency of the input speech detected by the emergency detecting means 1.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a technique for detecting urgency of utterance and changing recognition processing according to the urgency,
In particular, voice recognition applications (machines, vehicles, etc.) that include urgent instructions such as emergency stop among voice instructions.
And these simulation systems, games, etc.)
The present invention relates to a voice recognition device used in.

[0002]

2. Description of the Related Art In practical use of a voice recognition device, utterance variation is often a problem. Known utterance variations include, for example, temporal changes in voice, utterance variations under noise (Lombard effect), variations due to differences in utterance styles (isolated utterance words and continuous utterance words), and speech during utterance. There are fluctuations due to differences in emotions of the person. Due to such utterance variation, a difference occurs between the voice at the time of registration and the voice at the time of actual recognition, which is a problem in performing voice recognition processing.

Therefore, considering a voice recognition application in which an urgent instruction such as an emergency stop and an instruction only in normal time are mixed among voice instructions are considered.
Since an urgent voice instruction is given in an emergency, it becomes louder, louder, or faster than a voice registered in normal times. Therefore, the voice instruction in an emergency has a lower recognition rate than the voice instruction in the normal state.

Although it is considered that the candidates to be recognized are different in an emergency and in a normal situation, in the conventional voice recognition device, it is determined whether or not the input voice is urgent. Since there is no function, it has not been possible to change the candidate to be recognized in an emergency and in normal times, to make it easier to recognize any candidate, or to make it difficult to recognize any candidate.

In an emergency, since misrecognition and reject often give similar results, it is necessary to stop the reject and increase the percentage of correct answers = the number of correct answers / the number of trials. However, in the conventional voice recognition device, when the reliability of the first candidate for pattern matching is low, the recognition rate is increased to the number of correct answers / (the number of trials-the number of rejects) even if the rejection is performed and the correct response rate is decreased. Therefore, it was not possible to increase the correct answer rate in case of emergency.

Further, since particularly accurate recognition is required for an urgent voice instruction, in the conventional case, especially when the normal time and the emergency are mixed and used, the It was generally accepted that voice recognition is not suitable for some voice instructions.

As a conventional technique that solves such a problem and improves the recognition rate of urgent voices, there is, for example, a "voice recognition device" disclosed in Japanese Patent Publication No. 2-30039. This is the input voice pattern and the importance (urgency)
Of the specific registered voice pattern with high importance and other registered voice patterns of low importance, and by weighting the similarity between the specific registered voice pattern of high importance and the input voice pattern, The recognition rate for high-quality speech is increased.

[0008]

However, since voices of high importance are more likely to be selected as first place candidates, other voices are less likely to be selected as first place candidates, and normal voice instructions of low importance are given. Recognition rate is low. Further, since the number of important voice instructions is usually smaller than that of the other voice instructions, the recognition rate for almost all words will be low. Furthermore, since there is no distinction between an emergency and normal times, the above-mentioned problems always exist.

[0009]

According to a first aspect of the present invention, there is provided urgency detection means for detecting urgency from input voice, and input voice pattern generation means for analyzing the input voice to generate an input voice pattern. And a registered voice standard pattern storage means for storing a plurality of registered voice standard patterns registered in advance, and for each of these registered voice standard patterns by matching the input voice pattern with the registered voice standard pattern. Providing matching means for determining the distance between patterns of the input voice pattern,
A recognition determination means for determining the recognition result based on the distance between the patterns is provided.

According to a second aspect of the invention, in the first aspect of the invention, the input voice pattern generation means corrects the input voice pattern in accordance with the urgency detected by the urgency detection means.

According to a third aspect of the present invention, in the first aspect of the present invention, the registered voice standard pattern storage means also includes urgency information indicating whether or not each registered voice standard pattern is urgent. The matching means performs matching in correspondence with the urgency detected by the urgency detection means and the urgency information stored in the registered voice standard pattern storage means.

According to a fourth aspect of the present invention, in the invention of the first aspect, the recognition determining means determines the recognition result by changing the reject condition in accordance with the urgency detected by the urgency detecting means.

According to the invention of claim 5, claims 1, 2,
In the invention described in 3 or 4, the urgency detection means detects the power of voice as a parameter indicating the urgency.

According to the invention of claim 6, claims 1, 2,
In the invention described in 3 or 4, the urgency detection means detects the pitch of the voice as a parameter indicating the urgency.

According to the invention of claim 7, claims 1, 2,
In the invention described in 3 or 4, the urgency detection means detects the utterance speed of the voice as a parameter indicating the urgency.

According to the invention described in claim 8, claims 1, 2,
In the invention described in 3, 4, 5, 6 or 7, the urgency detection means uses a plurality of parameters including the power of the voice, the pitch of the voice and the utterance speed of the voice as parameters indicating the urgency.

[0017]

According to the first aspect of the invention, the voice recognition processing is performed in response to the urgency of the input voice detected by the urgency detecting means, thereby correcting the input voice pattern and changing the utterance in an emergency. It is possible to prevent a decrease in recognition rate due to, moreover, it becomes possible to easily recognize an urgent voice in an emergency, and further, it is possible to reduce the rejection in an emergency and increase the correct answer rate, As a result, it becomes possible to improve the voice recognition performance in an emergency without impairing the voice recognition performance in normal times.

According to the second aspect of the invention, the urgency of the input voice is detected by the urgency detecting means, and the input voice pattern is corrected in accordance with the urgency of the input voice, thereby changing the utterance in an emergency. It is possible to prevent a reduction in the recognition rate due to, and thereby it is possible to improve the voice recognition performance in an emergency without impairing the voice recognition performance in normal times.

According to the third aspect of the invention, the urgency of the input voice detected by the urgency detection means and the urgency information of each registered voice standard pattern stored in the registered voice standard pattern storage means are included. Correspondingly, by finding the distance between patterns by the matching means, it becomes possible to easily recognize an urgent input voice in an emergency, and this makes it possible to recognize an emergency voice without impairing the voice recognition performance in normal times. It is possible to improve the recognition performance.

According to the fourth aspect of the present invention, by changing the condition for rejecting the recognition result in accordance with the urgency of the input voice, it is possible to reduce the reject in an emergency and increase the correct answer rate. As a result, it becomes possible to improve the voice recognition performance in an emergency without impairing the voice recognition performance in normal times.

According to the fifth aspect of the invention, by detecting the power of the voice, it is possible to accurately grasp the urgency.

According to the sixth aspect of the invention, the urgency can be accurately grasped by detecting the pitch of the voice.

According to the seventh aspect of the present invention, it is possible to accurately grasp the urgency by detecting the vocalization speed of the voice.

According to the present invention, a plurality of parameters including the power of the voice, the pitch of the voice, and the utterance speed of the voice are used in combination, so that the urgency can be more accurately determined as compared with the case of one parameter. It becomes possible to capture.

[0025]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A first embodiment of the present invention will be described with reference to FIG. First, the outline of the overall configuration of this embodiment will be described with reference to FIG. On the output side of a microphone (not shown) to which voice is input, the urgency detection unit 1 as urgency detection means is provided.
And an input voice pattern generation unit 2 as an input voice pattern generation means are connected in parallel. A matching unit 3 as a matching unit and a recognition determining unit 4 as a recognition determining unit are sequentially connected in series to the output side of the input voice pattern generation unit 2. Also, the matching unit 3
A registered voice standard pattern storage unit 5 as a registered voice standard pattern storage means is connected to the. Furthermore, the output side of the urgency detection unit 1 is connected in parallel to the input voice pattern generation unit 2, the matching unit 3, and the recognition determination unit 4. That is, AB, AC, and AD are connected.

In such a structure, first, the outline of the voice recognition processing performed in this embodiment will be described. The urgency detection unit 1 detects urgency from the input voice input from the microphone, and at the same time, the input voice pattern generation unit 2 analyzes the input voice to generate an input voice pattern. The matching unit 3 performs matching between this input voice pattern and a plurality of registered voice standard patterns registered in the registered voice standard pattern storage unit 5 in advance, so that the input voice pattern of each of these registered voice standard patterns is matched. Find the distance between patterns. Then, the recognition determination unit 4 determines the recognition result based on the inter-pattern distance.

Next, the above-mentioned voice recognition processing will be specifically described. First, the urgency detection unit 1 detects urgency from the input voice input from the microphone. Here, the input voice in an emergency becomes louder, louder, and faster than normal voice. Therefore, by detecting the power of the voice, the pitch of the voice, and the speaking speed of the voice, respectively. , It is possible to detect whether or not the input voice is urgent. In this embodiment, the power of voice is used as the parameter indicating the urgency. That is, when the average power of one input voice is equal to or higher than a certain threshold, it is urgent, and in other cases, it is not urgent. This makes it possible to accurately grasp the urgency.

Further, the input voice pattern generation unit 2 performs voice section detection processing and acoustic analysis processing on the input voice input, and generates an input voice pattern from the acoustic analysis data in the voice section. That is, using a known processing method, the voice section detection process is performed by comparing the power of the voice with a threshold value, and the acoustic analysis process is performed by the LPC cepstrum analysis. As a result, the parameters of the input voice pattern are the LPC cepstrum and voice power. However, when the urgency detection unit 1 detects urgency, the input voice pattern is corrected using a known correction method. That is, it is well known that a loud voice (emergency voice) causes distortion not only on the level but also on the frequency axis as compared with the voice in normal times, and the voice fluctuation under noise (Lombard). (Effect) corresponds to this. The distortion of the input voice pattern due to such utterance variation is corrected using a known correction method.

The known processing method used for the above-mentioned voice section detection processing and acoustic analysis processing is, for example,
"Digital audio processing", Furui, Tokai University Press, etc. Further, as to the well-known correction method for the above-mentioned voicing variation under noise, for example, “Performance evaluation of the correction method by formant movement for voicing deformation under noise”, Takizawa et al., Proceedings of ASJ, 1- 8-9, September 1990.

On the other hand, the registered voice standard pattern storage unit 5 stores a plurality of registered voice standard patterns registered in advance, and urgency information indicating whether or not each of these registered voice standard patterns is urgent. Are also stored.
However, the parameters of the registered voice standard pattern are the LPC cepstrum and the power of voice similarly to the parameters of the input voice pattern. Table 1 shows an example of word sets and urgency information of each word stored in the registered voice standard pattern storage unit 5.

[0031]

[Table 1]

As shown in Table 1, words of stop, right and left with an urgency of 1 may be urgent, and words of forward, backward, acceleration and deceleration with other urgency of 0 are urgent. It is set so that there is no possibility. In the example shown in Table 1, whether there is an urgency or not is expressed by a binary value of 1 or 0, but it can be similarly implemented by expressing it by a continuous amount (a real number from 0 to 1). .. In this case, it becomes possible to detect the urgency more accurately.

In the matching section 3, the well-known DP (dynamic programming) is performed between the input voice pattern obtained by the input voice pattern generation section 2 and the registered voice standard pattern stored in the registered voice standard pattern storage section 5. ) Matching is performed to obtain the pattern distance between each registered voice standard pattern and the input voice pattern. However, when urgency is detected by the urgency detection unit 1, a registered voice standard pattern indicating that there is urgency among the plurality of registered voice standard patterns stored in the registered voice standard pattern storage unit 5. With respect to, the inter-pattern distance is corrected by multiplying by a constant smaller than 1 so that the inter-pattern distance to the input voice pattern becomes smaller.

Further, the recognition determination section 4 determines the recognition result using the inter-pattern distance obtained by the matching section 3. That is, when the ratio of the minimum inter-pattern distance to the second-smallest inter-pattern distance is equal to or greater than the reject threshold value greater than 1, the voice of the registered voice standard pattern given the minimum inter-pattern distance is set as the recognition result, and If it is not, reject it. However, when the urgency detection unit 1 detects the urgency, the reject threshold is set to be smaller or 1
To make it easier to recognize urgent voice.

As described above, the voice recognition processing is performed in response to the urgency of the input voice detected by the urgency detection unit 1 to correct the input voice pattern and recognize the recognition rate due to the utterance variation in the emergency. It is possible to prevent a decrease in the number of times, moreover, it is possible to easily recognize an urgent voice in an emergency, and further, in an emergency, it is possible to reduce the rejection and increase the correct answer rate. It is possible to improve the voice recognition performance in an emergency without impairing the voice recognition performance in normal times.

In this embodiment, the power of the voice is used as the parameter indicating the urgency, but the same can be done by using the pitch of the voice and the utterance speed of the voice.
Here, when the pitch of voice is used as the parameter indicating the urgency, the urgency is set if the average value of the pitches of one voice is equal to or more than a certain threshold value, and the urgency is not set otherwise. In this case, a known cepstrum analysis method is used to extract the pitch of the voice. In this way, it is possible to accurately grasp the urgency even by using the pitch of the voice.

When the voice utterance speed is used as a parameter indicating the urgency, the voice utterance speed is urgent if the voice utterance speed is equal to or higher than a certain threshold, and is not urgent otherwise. In this case, the speech production speed of the voice can be obtained from the peak of power per unit time and the dip. In this way, it is possible to accurately grasp the urgency even by using the vocalization speed of the voice. However, when the utterance speed is used as the parameter indicating the urgency, it is possible to improve the recognition of the urgent input voice by changing the restriction of the matching path of the matching unit 3.

Further, a parameter indicating the urgency of the above-mentioned voice power, voice pitch, and voice production speed,
Alternatively, the other parameters can be used together by weighting and then adding them. By using a plurality of parameters together in this way, it becomes possible to more accurately capture the urgency as compared with the case where there is one parameter.

Next, a second embodiment of the present invention will be described with reference to FIG. The same parts as those described in the first embodiment (see FIG. 1) are designated by the same reference numerals,
The explanation is also omitted. The configuration of this embodiment differs from the configuration of the first embodiment shown in FIG. 1 in the connection state. That is, as shown in FIG. 2, the output side of the urgency detection unit 1 is connected to the input voice pattern generation unit 2.
Also, the connection state is that only AB is connected,
As in the first embodiment shown in FIG. 1, AC and AD are not connected.

The voice recognition processing performed in this embodiment in such a configuration will be described with reference to FIG. First, the urgency detection unit 1 detects the urgency from the input voice input from a microphone (not shown) using the power of voice as a parameter indicating the urgency. That is, when the average power of one input voice is equal to or higher than a certain threshold, it is urgent, and in other cases, it is not urgent. In addition, the input voice pattern generation unit 2 uses the LPC cepstrum and the power of voice as parameters of the input voice pattern, and performs the voice segment detection process and the acoustic analysis process on the input voice to obtain the sound in the voice segment. An input voice pattern is generated from the analysis data. However, when the urgency detection unit 1 detects the urgency, the input voice pattern is corrected. On the other hand, the registered voice standard pattern storage unit 5 stores a plurality of registered voice standard patterns registered in advance. The parameters of this registered voice standard pattern are
The LPC cepstrum and the voice power as well as the parameters of the input voice pattern. And the matching unit 3
By this, well-known DP matching is performed between the input voice pattern obtained by the input voice pattern generation unit 2 and the registered voice standard pattern stored in the registered voice standard pattern storage unit 5, and each registered voice standard pattern The inter-pattern distance from the input voice pattern is calculated. Furthermore, the recognition determination unit 4 determines the recognition result using the inter-pattern distance obtained by the matching unit 3. That is, the voice of the registered voice standard pattern given the minimum inter-pattern distance is used as the recognition result, and the other voices are rejected.

As described above, by correcting the input voice pattern in response to the urgency of the input voice, it is possible to prevent the recognition rate from being lowered due to the utterance variation in an emergency. It is possible to improve the voice recognition performance in an emergency without impairing the voice recognition performance of.

In this embodiment, the power of voice is used as the parameter indicating the urgency of the urgency detecting section 1.
It is also possible to use the voice pitch and the voice production speed. Furthermore, the power of the voice, the pitch of the voice,
A parameter that indicates the urgency of the speaking rate of the voice, or
It is also possible to use them by adding other parameters after weighting them.

Next, a third embodiment of the present invention will be described with reference to FIG. The same parts as those described in the first embodiment (see FIG. 1) are designated by the same reference numerals,
The explanation is also omitted. The configuration of this embodiment differs from the configuration of the first embodiment shown in FIG. 1 in the connection state. That is, as shown in FIG. 3, the output side of the urgency detection unit 1 is connected to the matching unit 3. In addition, the connection state is such that AC is connected, and AB and AD are not connected as in the first embodiment shown in FIG.

The voice recognition processing performed in this embodiment in such a configuration will be described with reference to FIG. First, the urgency detection unit 1 detects the urgency from the input voice input from the microphone by using the power of voice as a parameter indicating the urgency. That is, when the average power of one input voice is equal to or higher than a certain threshold, it is urgent, and in other cases, it is not urgent. In addition, the input voice pattern generation unit 2 uses the LPC cepstrum and the power of voice as parameters of the input voice pattern to perform voice segment detection processing and acoustic analysis processing on the input voice, thereby An input voice pattern is generated from the acoustic analysis data. On the other hand, the registered voice standard pattern storage unit 5 stores a plurality of registered voice standard patterns registered in advance, and also includes urgency information indicating whether or not each of these registered voice standard patterns is urgent. Remember. The parameters of this registered voice standard pattern are the LPC cepstrum and the power of the voice, like the parameters of the input voice pattern. Then, the matching unit 3 performs well-known DP matching between the input voice pattern obtained by the input voice pattern generation unit 2 and the registered voice standard pattern stored in the registered voice standard pattern storage unit 5, The inter-pattern distance between the registered voice standard pattern and the input voice pattern is calculated. However, when urgency is detected by the urgency detection unit 1, a registered voice standard pattern indicating that there is urgency among the plurality of registered voice standard patterns stored in the registered voice standard pattern storage unit 5. With respect to, the inter-pattern distance is corrected by multiplying by a constant smaller than 1 so that the inter-pattern distance to the input voice pattern becomes smaller. This makes it easier to recognize an urgent input voice. Furthermore, the recognition determination unit 4 determines the recognition result using the inter-pattern distance obtained by the matching unit 3. That is, the voice of the registered voice standard pattern given the minimum inter-pattern distance is used as the recognition result, and the other voices are rejected.

As described above, since the inter-pattern distance is obtained in correspondence with the urgency of the input voice and the urgency information of each registered voice standard pattern, it is easy to recognize the urgent input voice in an emergency. This makes it possible to improve the voice recognition performance in an emergency without impairing the voice recognition performance in normal times.

In this embodiment, the power of voice is used as the parameter indicating the urgency of the urgency detecting section 1.
It is also possible to use the voice pitch and the voice production speed. Furthermore, the power of the voice, the pitch of the voice,
A parameter that indicates the urgency of the speaking rate of the voice, or
It is also possible to use them by adding other parameters after weighting them.

Next, a fourth embodiment of the present invention will be described with reference to FIG. The same parts as those described in the first embodiment (see FIG. 1) are designated by the same reference numerals,
The explanation is also omitted. The configuration of this embodiment differs from the configuration of the first embodiment shown in FIG. 1 in the connection state. That is, as shown in FIG. 4, the output side of the urgency detection unit 1 is connected to the recognition determination unit 4. In addition, the connection state is such that AD is connected, and AB and AC are not connected as in the first embodiment shown in FIG.

The voice recognition processing performed in this embodiment in such a configuration will be described with reference to FIG. First, the urgency detection unit 1 detects the urgency from the input voice input from the microphone by using the power of voice as a parameter indicating the urgency. That is, when the average power of one input voice is equal to or higher than a certain threshold, it is urgent, and in other cases, it is not urgent. In addition, the input voice pattern generation unit 2 uses the LPC cepstrum and the power of voice as parameters of the input voice pattern to perform voice segment detection processing and acoustic analysis processing on the input voice, thereby An input voice pattern is generated from the acoustic analysis data. On the other hand, the registered voice standard pattern storage unit 5 stores a plurality of registered voice standard patterns registered in advance. The parameters of this registered voice standard pattern are the LPC cepstrum and the power of the voice, like the parameters of the input voice pattern. And
The matching unit 3 performs well-known DP matching between the input voice pattern obtained by the input voice pattern generation unit 2 and the registered voice standard pattern stored in the registered voice standard pattern storage unit 5 to obtain each registered voice. The inter-pattern distance between the standard pattern and the input voice pattern is calculated. Furthermore, the recognition determination unit 4 determines the recognition result using the inter-pattern distance obtained by the matching unit 3. That is, the voice of the registered voice standard pattern given the minimum inter-pattern distance is used as the recognition result, and the other voices are rejected.
However, if the urgency detection unit 1 detects urgency,
By reducing the reject threshold or making it equal to 1 to make it more difficult to reject urgent input speech,
It is possible to improve the recognition rate of the input voice in an emergency.

As described above, by changing the rejection condition of the recognition determination section 4 in accordance with the urgency of the input voice, it is possible to reduce the rejection and increase the correct answer rate in an emergency. It is possible to improve the voice recognition performance in an emergency without impairing the voice recognition performance in normal times.

In the present embodiment, the power of voice is used as the parameter indicating the urgency of the urgency detector 1.
It is also possible to use the voice pitch and the voice production speed. Furthermore, the power of the voice, the pitch of the voice,
A parameter that indicates the urgency of the speaking rate of the voice, or
It is also possible to use them by adding other parameters after weighting them.

[0051]

According to the present invention, the urgency detecting means for detecting urgency from the input voice is provided, and the input voice pattern generating means for analyzing the input voice to generate the input voice pattern is provided in advance. A registered voice standard pattern storage means for storing a plurality of registered voice standard patterns is provided, and the input voice pattern for each of the registered voice standard patterns is obtained by matching the input voice pattern with the registered voice standard pattern. Since the matching means for determining the inter-pattern distance is provided and the recognition determination means for determining the recognition result based on the inter-pattern distance is provided, a voice corresponding to the urgency of the input voice detected by the urgency detection means is provided. By performing the recognition process, it is possible to correct the input voice pattern and prevent the recognition rate from decreasing due to voice fluctuation in an emergency. In addition, it is possible to easily recognize urgent voices in an emergency, and it is possible to increase the correct answer rate by reducing rejects in an emergency. It is possible to improve the voice recognition performance in an emergency without impairing it.

According to a second aspect of the present invention, in the first aspect of the invention, the input voice pattern generating means corrects the input voice pattern in accordance with the urgency detected by the urgency detecting means. It is possible to prevent a decrease in the recognition rate due to the utterance variation, and thereby to improve the voice recognition performance in an emergency without impairing the voice recognition performance in normal times.

According to a third aspect of the present invention, in the first aspect of the present invention, the registered voice standard pattern storage means also includes urgency information indicating whether or not each registered voice standard pattern is urgent. Since the matching means matches the urgency detected by the urgency detection means and the urgency information stored in the registered voice standard pattern storage means, By obtaining the inter-pattern distance by the matching means in correspondence with the urgency information of each registered voice standard pattern, it is possible to easily recognize an urgent input voice in an emergency. The voice recognition performance in an emergency can be improved without impairing the voice recognition performance.

According to a fourth aspect of the present invention, in the first aspect of the invention, the recognition determining means determines the recognition result by changing the reject condition corresponding to the urgency detected by the urgency detecting means. In an emergency, the rejection rate can be reduced and the correct answer rate can be increased, whereby the voice recognition performance in an emergency can be improved without impairing the voice recognition performance in normal times.

The invention according to claim 5 is the invention as claimed in claims 1, 2, and 3.
Or in the invention described in 4, since the urgency detection means detects the power of voice as a parameter indicating the urgency,
It is possible to accurately grasp the urgency.

The invention according to claim 6 is the same as claims 1, 2, and 3.
Or in the invention described in 4, since the urgency detection means detects the pitch of the voice as a parameter indicating the urgency,
It is possible to accurately grasp the urgency.

The invention according to claim 7 is the invention according to claims 1, 2, and 3.
Alternatively, in the invention described in item 4, the urgency detection means detects the utterance speed of the voice as a parameter indicating the urgency, so that the urgency can be accurately grasped.

The invention described in claim 8 is the same as in claim 1,
In the invention described in 3, 4, 5, 6 or 7, since the urgency detection means uses a plurality of parameters including the power of the voice, the pitch of the voice, and the utterance speed of the voice as the parameter indicating the urgency, The urgency can be grasped more accurately than in the case of one.

[Brief description of drawings]

FIG. 1 is a block circuit diagram showing a first embodiment of the present invention.

FIG. 2 is a block circuit diagram showing a second embodiment of the present invention.

FIG. 3 is a block circuit diagram showing a third embodiment of the present invention.

FIG. 4 is a block circuit diagram showing a fourth embodiment of the present invention.

[Explanation of symbols]

 1 urgency detection means 2 input voice pattern generation means 3 matching means 4 recognition determination means 5 registered voice standard pattern storage means

Claims (8)

[Claims] 1. An urgent detection means for detecting urgency from an input voice, an input voice pattern generation means for analyzing the input voice to generate an input voice pattern, and a plurality of registered voice standard patterns registered in advance. Registered voice standard pattern storage means for storing, matching means for obtaining the inter-pattern distance of the input voice pattern with respect to each of these registered voice standard patterns by performing matching between the input voice pattern and the registered voice standard pattern, A voice recognition device comprising a recognition determination means for determining a recognition result based on a distance between patterns. 2. The voice recognition device according to claim 1, wherein the input voice pattern generation means corrects the input voice pattern in response to the urgency detected by the urgency detection means. 3. The registered voice standard pattern storage means also stores urgency information indicating whether or not each registered registered voice standard pattern has an urgency, and the matching means detects the urgency detection means. 2. The voice recognition device according to claim 1, wherein matching is performed in correspondence with the urgency that has been recorded and the urgency information stored in the registered voice standard pattern storage means. 4. The voice recognition device according to claim 1, wherein the recognition determining means determines the recognition result by changing the reject condition in accordance with the urgency detected by the urgency detecting means. 5. The voice recognition device according to claim 1, 2, 3 or 4, wherein the urgency detection means detects the power of voice as a parameter indicating urgency. 6. The voice recognition device according to claim 1, 2, 3 or 4, wherein the urgency detecting means detects a pitch of voice as a parameter indicating urgency. 7. The voice recognition device according to claim 1, wherein the urgency detection means detects the utterance speed of the voice as a parameter indicating the urgency. 8. The urgentness detecting means uses a plurality of parameters including a power of voice, a pitch of voice, and a utterance speed of voice as a parameter indicating the urgency in combination. , 4, 5, 6 or 7 voice recognition device.