JP2014063018A - Extremely quiet voice input device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a voice input device having means for simple determination and removal of breath sound in an area where the pressure sound of an utterance tone that is neither noticed nor used is lower than that of the breath sound.SOLUTION: An extremely quiet voice input device according to the invention comprises breath sound removal means configured to remove breath sound after determining whether voice is the breath sound or extremely quiet voice based on a sound pressure threshold set between the breath sound and extremely quiet voice lower in sound pressure than the breath sound.

Description

The present invention relates to an apparatus for inputting voice instead of data input using a keyboard or the like.

Conventionally, when inputting a voice, an utterance sound is input from a microphone, but as usual, the utterance sound has a sound pressure level (about 60 dB) that is normally heard, and noise from the surroundings. Since the sound pressure level is smaller than this, there is no problem. However, although it is not a problem when working alone in a private room, it is extremely noisy for many workers to speak in the same room, and the speech of others is likely to be noisy. Therefore, an attempt to lower the sound pressure of the uttered sound is necessary. When the sound pressure of the uttered sound is lowered, the sound pressure approaches that of a breathing sound that was not noticed until now. The breathing sound is inaudible to the other person who is a little far away (neighboring) because the person who is breathing does not feel that he or she does not want to hear it. In such a state, voice input can be performed without being influenced by others, but since a breathing sound of the same level as the sound pressure of the uttered sound is picked up by the microphone, it is necessary to determine and remove the breathing sound. If the sound pressure is comparable, it cannot be divided by sound pressure, so a special means is required. Although not intended to remove breathing sounds, Patent Document 1 discloses a technique for calculating a fractal dimension number as a method for distinguishing between non-speech sounds and utterance sounds. Further, Patent Document 2 discloses an apparatus for discriminating noise from surroundings and speech sounds, performing chaos analysis for the purpose of removing surrounding noises, and calculating a Lyapunov exponent. Both perform calculation processing based on mathematical theory, but the calculation processing is heavy.
Note that the above-described voice input is performed in the mouth of the microphone or, in a special case, in the oral cavity, as normally considered. Although this is a completely different field, there is what is commonly called a NAM microphone. This is to capture the sound generated in the back of the throat with a microphone in contact with the skin under the ears. Since these sounds are not received, these minute sounds are not within the scope of the present application.

JP2007-264567 JP 2003-99094 A

The problem of the present invention is that a simple breathing sound in a region where the sound pressure of an uttered sound, which has not been noticed and used because it is noisy more than a necessary signal, is smaller than that of a breathing sound (very small sound region) An object of the present invention is to provide a voice input device having means for discrimination and removal.

The very small sound input device of the present invention determines whether the sound is a breathing sound or a very small sound based on a sound pressure threshold value set between a breathing sound and a very small sound having a sound pressure smaller than this, A breathing sound is removed and will be described in detail below.

The invention described in claim 1 is a very small voice input device, which is a voice input device that cannot be heard by another person even when a large number of speakers perform voice input, and does not hinder the input of others.
In order to acquire a sound including a very small sound emitted by a speaker having a sound pressure smaller than that of a breathing sound, a microphone installed in a closed closed section around the mouth or in the mouth or around the mouth and the sound acquired by the microphone are used. By comparing the analog-to-digital conversion AD converter, the memory for storing the audio data output from the AD converter, and the sound pressure threshold value for comparison and determination set between the breathing sound and the micro-sound When the sound pressure of the sound is larger than the sound pressure threshold, the sound is removed as the breathing sound, and when the sound pressure of the sound is smaller than the sound pressure threshold, the sound is very small. It has a breathing sound removing means that is left as speech, an AD converter, the memory, and a central processing unit that controls the breathing sound removing means.

The invention described in claim 2 is the micro audio input device according to claim 1,
When the sound pressure of the voice is larger than the sound pressure threshold, the breathing sound removing means is set between the breathing sound and a normal sound pressure (60 dB) that is a sound pressure of a normal utterance normally made by a speaker. When the sound pressure of the sound is smaller than the second sound pressure threshold by comparing with a second sound pressure threshold for comparison determination, the sound is removed as the breathing sound, and the sound is removed. When the sound pressure is higher than the second sound pressure threshold, the voice is removed as the normal utterance, and a determination is made to instruct the voice input again with the very small voice.

The invention described in claim 3 is the micro audio input device according to claim 1 or 2,
The breathing sound removing means or the central processing unit acquires the voice data from the memory as a group with a time interval smaller than a unit of one time interval of the breathing sound or inspiration of the breathing sound, When adjacent groups are continuous as exhalation, inspiration, or speech, they are the same group, and when they are not continuous in one group, they are divided into different groups, and the same processing is performed for the same group.

A fourth aspect of the present invention is the micro audio input device according to any one of the first to third aspects,
In order to switch between the case where the speaker sings the normal utterance alone and the case where the utterance is made with the very minute voice, the sensitivity of the microphone is low and the breathing sound removing means does not operate. Input mode switching means for switching between the normal speech corresponding mode and the microphone having high sensitivity and switching between the ultra-speech speech mode in which the breathing sound removing means operates.

Since it is configured as described above, in the micro audio input device of the present invention, even if a large number of workers perform audio input, it cannot be heard by others, does not become noise of other people's input, and eliminates breathing sounds. This is convenient for monitoring and enables processing convenient for data processing.

It is a figure which shows one embodiment of a structure and process flow of the micro audio | voice input apparatus of this invention. It is a figure which shows the sound pressure level of the speech sound of an audio | voice input, and a breathing sound. It is a figure which shows the time transition of the audio | voice input input into the micro audio | voice input apparatus of this invention, and the speech sound pressure (intensity) after a breathing sound removal.

The very small voice input device of the present invention is a voice input device that cannot be heard by others even when many speakers perform voice input, and does not hinder other people's input,
In order to acquire sound including extremely small sound emitted by a speaker with a sound pressure lower than that of breathing sound, a microphone placed in a closed closed section around the mouth or mouth or around the mouth and the sound acquired by the microphone are analog to digital By comparing the AD converter to be converted, the memory for storing the audio data output from the AD converter, and the sound pressure threshold value for comparison and determination set between the breathing sound and the extremely small sound, the sound pressure of the sound If the sound pressure is greater than the sound pressure threshold, the sound is removed as a breathing sound, and if the sound pressure of the sound is smaller than the sound pressure threshold, the breathing sound removing means that leaves the sound as a very small speech, An AD converter, a memory, and a central processing unit for controlling the breathing sound removing means; Hereinafter, it demonstrates along an Example.

FIG. 2 is a diagram illustrating the sound pressure levels of speech sounds and breathing sounds input by voice. A microphone (especially convenient, a unidirectional microphone) is placed in the speaker's mouth or, in special cases, in the mouth or in a closed section around the mouth and speaks. As shown in 2-B, the utterance sound pressure of the normal conversation is about 60 dB. On the other hand, the breathing sound pressure when a human is breathing is said to be 10 to 20 dB. In the speech pressure of normal conversation, since the respiratory sound pressure is extremely small, it is not worrisome, and the discrimination by sound pressure is easy.
When the sound pressure is gradually reduced from the sound pressure of normal conversation (decreasing to a lower voice), the speech sound pressure is almost the same as or slightly larger than the breathing sound as shown as the minute speech sound, and only the sound pressure It becomes difficult to distinguish between the two. The distinction here is conventionally performed by means of a mathematical approach such as chaos analysis, but it has been described in the conventional example that the calculation process is complicated and heavy.
As a means for simplifying the processing, it can be proposed to make a distinction based on a difference in characteristics including fluctuations in the air flow of speech sound and breathing sound.

When the sound pressure is further lowered from this minute utterance sound, it is smaller than the breathing sound pressure as described as a very minute utterance sound, so naturally it cannot be heard by another person's ear next to it. Moreover, although it is smaller than the breathing sound, an utterance sound can be acquired from the microphone. However, since the breathing sound is louder than the desired utterance sound, when monitoring the input content, the breathing sound cannot be heard. As data processing, breathing sounds are unnecessary, so I want to remove them. Although this sound pressure range is usually out of common sense, it was not used, but even if it is performed by many people, if you stick to quiet voice input that does not disturb others, you can enjoy the advantage that it will not be heard by others. it can. However, there are weaknesses.
If you happen to have a big utterance from someone else, you also need to remove it.
The present application is a device that performs voice input for extremely small sound utterances.
2-A shows the above as a table.

FIG. 3 is a diagram showing the time transition of the voice input and the speech sound pressure (intensity) after removing the breathing sound input to the very small voice input device of the present invention. In 3-A and 3-B, the horizontal axis represents time, and the vertical axis conceptually represents the change in voice intensity (not the actual waveform, but the range in which the waveform exists). 3-A corresponds to after voice input, and 3-B corresponds to after desired breathing sound removal. In 3-A, when there is inspiration and expiration with intensity greater than the intensity of the utterance sound, there is no utterance, inspiration and expiration are repeated. When there is an utterance, the utterance is normally made at the place of exhalation, and the utterance is made with a longer duration than in the case of exhalation. It is well known that utterances vary in length depending on the content of the spoken language, and additional utterances are often made after breathing.

FIG. 1 is a diagram showing an embodiment of a configuration and a processing flow of a micro audio input device of the present invention. 1-A shows an example of the hardware configuration, and 1-B shows an example of the processing flow of the breathing sound removing means. 1-A includes a microphone 101, an ADC (analog / digital converter) 102, a CPU (central processing unit) 103, a memory 104, and a breathing sound removing unit 105. Note that the breathing sound removing means 105 performs determination and processing, so when processing by software, the CPU 103 and the memory 104 (may include a hard disk).
It is possible to think that it is inherent in

Since the microphone 101 acquires a very small sound smaller than the breathing sound, the microphone 101 is installed in a closed section closed in the mouth or in a special case in the mouth or around the mouth. Also, a unidirectional microphone is advantageous so as not to pick up sound from the surroundings. In the closed section closed in the mouth or around the mouth, a unidirectional microphone can be used.
Analog audio data acquired by the microphone 101 is input to the ADC 102, converted into a digital quantity, and stored in the memory 104. The CPU 103 controls the operation of the ADC 102 and the storage operation of the memory 104. The breathing sound removing means 105 reads data for each desired time interval from the memory 104, performs judgment and processing, and stores the data from which the breathing sound has been removed in the memory 104 (or other memory) or other places not described. Send data to.

The data reading time interval is preferably a time interval shorter than that of the breathing time interval (about 1 to 2 seconds). For example, so that the sampling theorem is satisfied with respect to the audible sound range (˜20 kHz), the data acquisition at the ADC is performed in an audible sound range such as 50 kHz so as to satisfy the sampling theorem with respect to the audible sound range (˜20 kHz). Performing at twice or more acquisition frequency (sampling frequency), but to identify a group of one inspiration, exhalation, and utterance, collect the acquired data at a reading time interval of about 1-2 seconds or less I do. The data read at each time interval means that continuous sound pressure time intervals are grouped as belonging to the same sound as necessary, and are subjected to the same processing. This is because the breathing sounds are also inhaled and exhaled, separated by a silent interval, and each may be treated as one group. Also, the uttered sound is emitted in the exhalation section and either a plurality of words are continuous or one word is separated by a silent section. If they are continuous, the same processing can be performed for the section as the same group. Moreover, when it does not continue within one group, it can be divided into different groups.

It should be noted that the grouping and the processing can be determined in consideration of the following utterance characteristics.
<Sound characteristics>
・ Speech is often from the beginning of exhalation (air flow is outward). There is very little beginning in the middle of exhalation unless intentional.
When the first pattern of exhalation appears, it can be determined that there is a breathing sound from there to the interval between exhalation and the next inspiration. If the first pattern of exhalation does not appear and the utterance starts, the utterance section is entered.
・ Do not speak in inhalation (air flow is inward).
・ The breathing sound has one direction of air flow, and there is no fluctuation compared to speech.
・ The sound pressure of the breathing sound is less variable than the utterance sound because the same individual always breathes in the same way.
・ Do not speak with exhaled breath.
・ Utterances have many components above 500Hz. The breathing sound has many components of 500 Hz or less.

An example of the operation of the breathing sound removing unit 105 will be described in 1-B.
An utterance is input from the microphone 101. After the speech input, it is determined whether or not the sound pressure is larger than the threshold value 1 through the storage of data in the memory 104 or directly. The threshold value 1 is set to a sound pressure value between the breathing sound and the extremely small sound. For example, referring to 2-B in FIG. 2, a periphery of 10 dB is appropriate. This value may be set to an appropriate value between the breathing sound and the utterance sound by actually measuring the measured value of the voice input in FIG. When the sound pressure is smaller than the threshold value 1, it is determined as a very small utterance sound and data is not removed and continues. When the sound pressure is larger than the threshold value 1, 2) It is determined whether the sound pressure is larger than the threshold value 2. The threshold 2 is set to a sound pressure value between the breathing sound and the normal utterance sound. For example, 30 dB is appropriate.
When the sound pressure is smaller than the threshold value 2, it is determined as a breathing sound and removed, and this time interval is silenced. If the sound pressure is greater than the threshold value 2, it is determined that a normal speech sound has been mixed and removed, and an instruction “Please input again at an appropriate level” is issued. This determination process is repeated for voice input as time progresses.

Note that the processing after the determination of whether the sound pressure is greater than the threshold value 2 is not essential, but in reality, it may be accidentally input with a loud voice or a loud voice of another person may be picked up. It is convenient to have such a determination process. In addition, in the case of a large number of people, such breathing sound removing means is operated. However, in the case of a single work in a room, it may be input with the sound pressure of a normal utterance. It is convenient to provide an input mode switching means for switching between the two because it is sufficient to input with lowering the sensitivity of the microphone without worrying about the operation / non-operation of the removing means.
Through the processing as described above, the voice input of 3-A in FIG. 3 can obtain voice data after removing the breathing sound of 3-B.

Since it is configured as described above, the very small voice input device of the present invention can not hear other people even if many workers input voice, does not become noise of other people's input, and eliminates breathing sounds. Since it is convenient for monitoring and is provided for data processing, the industrial applicability is extremely large.

101 Microphone 102 ADC
103 CPU
104 memory 105 breathing sound removal means

Claims (4)

A voice input device that cannot be heard by other people even when a large number of speakers input voice, and does not interfere with the input of others.
In order to acquire a sound including a very small sound emitted by a speaker having a sound pressure smaller than that of a breathing sound, a microphone installed in a closed closed section around the mouth or in the mouth or around the mouth and the sound acquired by the microphone are used. By comparing the analog-to-digital conversion AD converter, the memory for storing the audio data output from the AD converter, and the sound pressure threshold value for comparison and determination set between the breathing sound and the micro-sound When the sound pressure of the sound is larger than the sound pressure threshold, the sound is removed as the breathing sound, and when the sound pressure of the sound is smaller than the sound pressure threshold, the sound is very small. A very small voice input device comprising: a breathing sound removing unit that is left as speech, a central processing unit that controls the AD converter, the memory, and the breathing sound removing unit When the sound pressure of the voice is larger than the sound pressure threshold, the breathing sound removing means is set between the breathing sound and a normal sound pressure (60 dB) that is a sound pressure of a normal utterance normally made by a speaker. When the sound pressure of the sound is smaller than the second sound pressure threshold by comparing with a second sound pressure threshold for comparison determination, the sound is removed as the breathing sound, and the sound is removed. When the sound pressure is larger than the second sound pressure threshold, the voice is removed as the normal utterance, and a determination is made to instruct the voice input again with the very small voice. Item 2. The micro audio input device according to Item 1. The breathing sound removing means or the central processing unit acquires the voice data from the memory as a group with a time interval smaller than a unit of one time interval of the breathing sound or inspiration of the breathing sound, When adjacent groups are continuous as exhalation, inspiration, or utterance, the same group is used, and when they are not continuous in one group, they are divided into different groups, and the same processing is performed for each same group. Item 3. The minute voice input device according to item 1 or 2. In order to switch between the case where the speaker sings the normal utterance alone and the case where the utterance is made with the very minute voice, the sensitivity of the microphone is low and the breathing sound removing means does not operate. 4. An input mode switching means for switching between the normal speech corresponding mode and the microphone having a high sensitivity, and switching between the micro-speech speech modes in which the breathing sound removing means is activated. The micro audio input device according to any one of the above.

Images