Sound output method, sound output device, and sound output control program

Abstract

To accurately remove noise unnecessary for a worker.SOLUTION: The sound output apparatus 1 includes an imaging unit 2 mounted on a head of an operator, a sound collecting unit 3 that detects a sound around the operator, a storage unit 4 that stores a plurality of predetermined first specific sounds, a determining unit 5 that determines relevance between imaging data detected by the imaging unit 2 and the first specific sounds, a noise predicting unit 6 that predicts an actual noise in accordance with a determination result of the determining unit 5, and an anti-noise generating unit 7 that generates a noise for noise cancellation that is a sound for noise reduction on the basis of the noise predicted by the noise predicting unit 6. And an output part 8 for outputting the generated noise for noise cancellation to an operator.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an audio output method, an audio output device, and an audio output control program. [Background technology]

[0002] For example, Patent Document 1 discloses a noise reduction device that reduces noise generated by a noise source, and that includes a characteristic input unit that inputs characteristics of noise previously measured at a location where a noise source is present, an ambient sound acquisition unit that acquires an ambient sound signal, which is an audio signal obtained by collecting ambient sound, and a noise reduction processing unit that attenuates signal components from the ambient sound signal that correspond to the characteristics of the input noise. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2012-93445 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the noise reduction device in Patent Document 1 simply removes the noise to be removed using the characteristics of that noise measured in the past. Therefore, in Patent Document 1, if the characteristics of the past noise do not match the characteristics of the current noise, there is a risk that the noise to be removed will not be removed. [Means for solving the problem]

[0005] The audio output of the present invention is characterized by storing a specific sound associated with the movement of a noise source around a worker, capturing an image of the noise source with an imaging unit worn by the worker, detecting sounds around the worker, detecting the movement of the noise source from the information obtained by capturing the image with the imaging unit, determining whether the detected movement of the noise source is the movement stored in association with the specific sound, and if it is determined that the detected movement of the noise source is the stored movement, generating a noise reduction sound based on the specific sound and outputting the noise reduction sound to the worker. [Effects of the Invention]

[0006] The present invention outputs to the worker a noise reduction sound that is generated by taking into consideration factors related to the current generation of noise, and therefore can accurately remove noise that is unnecessary for the worker. [Brief explanation of the drawings]

[0007] [Figure 1] FIG. 1 is an explanatory diagram showing a schematic system configuration of an audio output device according to a first embodiment of the present invention. [Figure 2] 4 is a flowchart showing an example of a control flow in the audio output device of the first embodiment. [Figure 3] FIG. 10 is an explanatory diagram showing a schematic system configuration of an audio output device according to a second embodiment of the present invention. [Figure 4] 10 is a flowchart showing an example of a control flow in the audio output device according to the second embodiment. [Figure 5] FIG. 10 is an explanatory diagram showing a schematic system configuration of an audio output device according to a third embodiment of the present invention. [Figure 6] 10 is a flowchart showing an example of a control flow in the audio output device according to the third embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0008] An embodiment of the present invention will be described in detail below with reference to the drawings.

[0009] FIG. 1 is an explanatory diagram that schematically shows the system configuration of an audio output device 1 of the first embodiment.

[0010] The audio output device 1 has an imaging unit 2 worn on the worker's head, a sound collection unit 3 that detects sounds around the worker, a memory unit 4 that stores a plurality of predetermined first specific sounds, which are specific sounds, a judgment unit 5 that judges the correlation between the imaging data detected by the imaging unit 2 and the first specific sounds, a noise prediction unit 6 that predicts actual noise based on the judgment result of the judgment unit 5, an anti-noise generation unit 7 that generates noise cancellation noise, which is noise reduction sound, based on the noise predicted by the noise prediction unit 6, and an output unit 8 that outputs the generated noise cancellation noise to the worker.

[0011] In the sound output device 1, the imaging unit 2, sound collection unit 3, and output unit 8 are located on the worker's side so that they can be worn by the worker. For example, the sound collection unit 3 and the output unit 8 may be arranged in a single component (e.g., earphones) worn by the worker.

[0012] On the other hand, in the audio output device 1, the memory unit 4, determination unit 5, noise prediction unit 6, and anti-noise generation unit 7 do not necessarily need to be located on the worker's side, and may be constructed, for example, in a server 9 provided at a location away from the worker and a noise source (described later). The server 9 is equipped with a CPU, ROM, RAM, and an input / output interface, and in other words, is a computer.

[0013] The imaging unit 2 is a wide-angle camera worn on the worker's head, and is fixed to, for example, the worker's helmet, etc., and is capable of capturing images of noise sources that emit noise in association with the operation of machinery and equipment around the worker, as well as the worker's movements. The imaging unit 2 also has a function of transmitting information about the noise source obtained by capturing the image (imaging information) to the server 9, which will be described later. Here, the information obtained by the imaging unit 2 is, for example, information in the form of images or videos.

[0014] The sound collection unit 3 is a microphone capable of collecting sounds around the worker, and is fixed to, for example, the worker's helmet, etc. The sound collection unit 3 also has a function of transmitting collected sound information around the worker to the server 9.

[0015] The storage unit 4 stores in advance a plurality of predetermined first specific sounds associated with the actions of a plurality of noise sources around the worker.

[0016] The first specific sound is associated (labeled) with the operation of the noise source using a predetermined learning model (first learning model) and stored in the storage unit 4. The first learning model is generated by, for example, machine learning.

[0017] More specifically, the first specific sound is stored in the storage unit 4 in association (labeled) with each operation of the noise source.

[0018] The determination unit 5 detects the movement of the noise source from the information obtained by capturing an image with the imaging unit 2, and determines whether the detected movement of the noise source is a movement stored in association with the first specific sound.

[0019] In more detail, the judgment unit 5 predicts the future behavior of the noise source from the detected behavior of the noise source, and judges whether the predicted future behavior of the noise source is a behavior stored in association with the first specific sound.

[0020] These determinations made by the determination unit 5 are made using a predetermined learning model (second learning model) that has been learned in advance. The second learning model is, for example, based on machine learning.

[0021] When it is determined that the detected movement of the noise source is a movement that is associated with a first specific sound and stored in advance, the noise prediction unit 6 generates noise that is predicted to be emitted from the noise source based on the associated first specific sound.

[0022] In more detail, the noise prediction unit 6 generates noise that is expected to be emitted from the noise source based on a first specific sound associated with the future behavior of the noise source predicted from the detected behavior of the noise source.

[0023] The anti-noise generation unit 7 generates noise for noise cancellation as a noise reduction sound that reduces the noise generated by the noise prediction unit 6. That is, the anti-noise generation unit 7 generates noise for noise cancellation based on a predetermined first specific sound associated with the detected operation of the noise source. In other words, the anti-noise generation unit 7 generates noise for noise cancellation in accordance with the determination result of the determination unit 5.

[0024] The anti-noise generation unit 7 has a function of transmitting the generated noise for noise cancellation to the output unit 8. In other words, the server 9 has a function of transmitting the noise for noise cancellation to the output unit 8.

[0025] The audio output device 1 of the first embodiment described above can set a correspondence between the operation of the vibration generating source and the noise for noise cancellation by storing the first specific sound in association with the operation of the vibration generating source using a learning model (first learning model).

[0026] In other words, the audio output device 1 outputs to the worker noise for noise cancellation that is generated taking into consideration factors related to the current generation of noise, and therefore can accurately remove noise that is unnecessary for the worker.

[0027] Furthermore, since the audio output device 1 can predict the noise emitted by the noise source, it can output noise cancellation noise to the worker in real time in response to the noise generated from the noise source, which also makes it possible to accurately remove noise that is unnecessary for the worker.

[0028] FIG. 2 is a flowchart showing an example of a control flow in the audio output device 1 of the first embodiment.

[0029] In step S11, the imaging unit 2 captures an image of a noise source and the like, and the sound collection unit 3 collects (detects) sounds around the worker.

[0030] In step S12, the future behavior of the noise source is predicted based on the imaging results from the imaging unit 2.

[0031] In step S13, the first specific sound associated with the predicted operation of the noise source is referenced.

[0032] In step S14, the referenced first specific sound and the sound collected by the sound collection unit 3 are input to predict the noise that is predicted to be generated in the future from the noise source.

[0033] In step S15, noise for noise cancellation is generated based on the noise predicted in step S14.

[0034] In step S16, the noise for noise cancellation is reproduced through the output unit 8.

[0035] Specifically, the first embodiment described above detects a work robot as a noise source in a factory or the like, detects the start point of the robot's repetitive work, predicts the average sound associated with the robot and the time it will finish, predicts in real time the noise that the robot will make during the repetitive work, and performs noise cancellation of the predicted noise that the robot will make in the future.

[0036] Another embodiment of the present invention will be described below, in which the same components as those in the above-described embodiment are designated by the same reference numerals and redundant description will be omitted.

[0037] A second embodiment of the present invention will be described with reference to Fig. 3. An audio output device 21 of the second embodiment has substantially the same configuration as the audio output device 1 of the first embodiment described above, but as shown in Fig. 3, an audio processing unit 22 is added.

[0038] The audio output device 21 of the second embodiment includes an imaging unit 2 worn on the worker's head, a sound collection unit 3 that detects sounds around the worker, a memory unit 4 that stores a plurality of predetermined high-frequency first specific sounds and low-frequency first specific sounds, an audio processing unit 22 that separates the sounds collected by the sound collection unit 3 into high-frequency sounds on the high-frequency side and low-frequency sounds on the low-frequency side, a judgment unit 5 that judges the correlation between the imaging data detected by the imaging unit 2 and the high-frequency first specific sounds and the low-frequency first specific sounds, a noise prediction unit 6 that predicts actual noise based on the judgment result of the judgment unit 5, an anti-noise generation unit 7 that generates noise cancellation noise, which is noise reduction sound, based on the noise predicted by the noise prediction unit 6, and an output unit 8 that outputs the generated noise cancellation noise to the worker.

[0039] In the audio output device 21, the memory unit 4, audio processing unit 22, determination unit 5, noise prediction unit 6, and anti-noise generation unit 7 do not necessarily need to be located on the worker side, but may be constructed, for example, in a server 9 provided at a location away from the worker and noise sources. The server 9 is equipped with a CPU, ROM, RAM, and an input / output interface, in other words, is a computer.

[0040] The sound processing unit 22 separates the sound collected by the sound collection unit 3 into predetermined high-frequency sound and predetermined low-frequency sound. The high-frequency sound is sound collected by the sound collection unit 3 with a frequency equal to or higher than a predetermined frequency threshold, for example, sound in the inaudible range. The low-frequency sound is sound collected by the sound collection unit 3 with a frequency lower than a predetermined frequency threshold, for example, sound in the audible range. The frequency threshold is, for example, 20 kHz.

[0041] The memory unit 4 stores a plurality of predetermined high-frequency first specific sounds associated with each of the actions of a plurality of noise sources around the worker, and a plurality of predetermined low-frequency first specific sounds associated with each of the actions of a plurality of noise sources around the worker.

[0042] The high-frequency first specific sound, which is a specific sound, is a high-frequency specific sound, and is associated (labeled) with the behavior of the noise generating source using a predetermined learning model (first learning model) and stored in the storage unit 4. More specifically, the high-frequency first specific sound is associated (labeled) individually with each behavior of the noise generating source and stored in the storage unit 4.

[0043] The low-frequency first specific sound, which is a specific sound, is a low-frequency specific sound, and is associated (labeled) with the behavior of the noise generating source using a predetermined learning model (first learning model) and stored in the storage unit 4. More specifically, the low-frequency first specific sound is associated (labeled) individually with each behavior of the noise generating source and stored in the storage unit 4.

[0044] That is, in the second embodiment, the specific sound is a high-frequency first specific sound which is a high-frequency specific sound and a low-frequency first specific sound which is a low-frequency specific sound.

[0045] The determination unit 5 detects the movement of the noise source from the information obtained by capturing an image with the imaging unit 2, and determines whether the detected movement of the noise source is a movement stored in association with the high-frequency first specific sound. More specifically, the determination unit 5 predicts the future movement of the noise source from the detected movement of the noise source, and determines whether the predicted future movement of the noise source is a movement stored in association with the high-frequency first specific sound.

[0046] The determination unit 5 detects the movement of the noise source from the information obtained by capturing an image with the imaging unit 2, and determines whether the detected movement of the noise source is a movement stored in association with the first low-frequency specific sound. More specifically, the determination unit 5 predicts the future movement of the noise source from the detected movement of the noise source, and determines whether the predicted future movement of the noise source is a movement stored in association with the first low-frequency specific sound.

[0047] These determinations made by the determination unit 5 are made using a predetermined learning model (second learning model) that has been learned in advance.

[0048] When it is determined that the detected behavior of the noise source is a behavior that is associated with the high-frequency first specific sound and stored in advance, the noise prediction unit 6 generates noise that is predicted to be emitted from the noise source based on the associated high-frequency first specific sound. More specifically, the noise prediction unit 6 generates noise that is predicted to be emitted from the noise source based on the high-frequency first specific sound that is associated with the future behavior of the noise source predicted from the detected behavior of the noise source.

[0049] Furthermore, when it is determined that the detected behavior of the noise source is a behavior associated with a low-frequency first specific sound and stored in advance, the noise prediction unit 6 generates noise predicted to be emitted from the noise source based on the associated low-frequency first specific sound. More specifically, the noise prediction unit 6 generates noise predicted to be emitted from the noise source based on the low-frequency first specific sound associated with the future behavior of the noise source predicted from the detected behavior of the noise source.

[0050] In addition, if there is no high-frequency first specific sound associated with the detected operation of the noise source, the noise prediction unit 6 classifies the separated high-frequency sound as environmental sound and mixes the separated high-frequency sound with the low-frequency first specific sound to generate noise predicted to be emitted from the noise source.

[0051] In other words, when it is determined that the detected behavior of the noise source is a behavior associated with a first specific sound and stored in advance, the noise prediction unit 6 generates noise predicted to be emitted from the noise source based on the associated first specific sound. More specifically, the noise prediction unit 6 generates noise predicted to be emitted from the noise source based on the first specific sound associated with the future behavior of the noise source predicted from the detected behavior of the noise source.

[0052] The anti-noise generation unit 7 generates noise for noise cancellation as a noise reduction sound that reduces the noise generated by the noise prediction unit 6. That is, the anti-noise generation unit 7 generates noise for noise cancellation based on the high-frequency first specific sound and the low-frequency first specific sound associated with the detected operation of the noise source. In other words, the anti-noise generation unit 7 generates noise for noise cancellation in accordance with the determination result of the determination unit 5.

[0053] The audio output device 21 of the second embodiment can achieve substantially the same effects as the audio output device 1 of the first embodiment described above, and can also accurately remove noise in the high frequency range that is unnecessary for workers.

[0054] FIG. 4 is a flowchart showing an example of a control flow in the audio output device 21 of the second embodiment.

[0055] In step S21, the imaging unit 2 captures an image of a noise source and the like, and the sound collection unit 3 collects (detects) sounds around the worker.

[0056] In step S22, the audio processing unit 22 separates the collected audio into a collected audio in the low frequency band and a collected audio in the high frequency band.

[0057] In step S23, the future behavior of the noise source is predicted based on the imaging results from the imaging unit 2.

[0058] In step S24, the low-frequency first specific sound associated with the predicted operation of the noise source is referenced.

[0059] In step S25, the high-frequency first specific sound associated with the predicted operation of the noise source is referenced.

[0060] In step S26, the referenced low-frequency first specific sound and the collected low-frequency sound are input to predict low-frequency noise that is predicted to be generated in the future from the noise source.

[0061] In step S27, the referenced high-frequency first specific sound and the collected high-frequency sound are input to predict high-frequency noise that is predicted to be generated in the future from the noise source.

[0062] In step S28, noise in the entire frequency range is predicted from the predicted low-frequency band noise and the predicted high-frequency band noise. That is, in step S28, the predicted low-frequency band noise and the predicted high-frequency band noise are mixed. If there is no high-frequency first specific sound associated with the detected operation of the noise source, the separated high-frequency audio is mixed with the low-frequency first specific sound after adjusting the sound pressure, etc., and predicted as noise in the entire frequency range.

[0063] In step S29, noise for noise cancellation is generated based on the noise predicted in step S28.

[0064] In step S30, the noise for noise cancellation is reproduced through the output unit 8.

[0065] A third embodiment of the present invention will be described with reference to Fig. 5. An audio output device 31 of the third embodiment has substantially the same configuration as the audio output device 1 of the first embodiment described above, but as shown in Fig. 5, a person detection unit 32 and a noise processing unit 33 are added.

[0066] The audio output device 31 of the third embodiment includes an imaging unit 2 worn on the head of a worker, a sound collection unit 3 that detects sounds around the worker, a memory unit 4 that stores information on a plurality of predetermined first specific sounds, a plurality of predetermined second specific sounds, and a plurality of pre-registered people, a determination unit 5 that determines the correlation between the imaging data detected by the imaging unit 2 and the first and second specific sounds, a noise prediction unit 6 that predicts actual noise based on the determination result of the determination unit 5, an anti-noise generation unit 7 that generates noise cancellation noise, which is noise reduction audio, based on the noise predicted by the noise prediction unit 6, a person detection unit 32 that detects whether a pre-registered person has been detected based on information captured by the imaging unit 2 and information in the memory unit 4, a noise processing unit 33 that captures external sounds and adds them to the noise cancellation noise when the distance between the pre-registered person and the worker is equal to or less than a predetermined distance (a predetermined value), and an output unit 8 that outputs the generated noise cancellation noise toward the worker. The external sounds are captured, for example, from the sound collection unit 3.

[0067] In the audio output device 31, the memory unit 4, judgment unit 5, person detection unit 32, noise prediction unit 6, anti-noise generation unit 7, and noise processing unit 33 do not necessarily need to be located on the worker's side, but may be constructed, for example, in a server 9 provided at a location away from the worker and noise sources. The server 9 is equipped with a CPU, ROM, RAM, and an input / output interface, in other words, a computer.

[0068] The storage unit 4 pre-stores a plurality of predetermined first specific sounds associated with each action of a noise source around the worker, a plurality of predetermined second specific sounds associated with each action of the worker, and person information that is information about a plurality of pre-registered people. In other words, the third embodiment considers the worker and the noise source as noise sources.

[0069] The first specific sound, which is a specific sound, is stored in the storage unit 4 in association (labeled) with the operation of the noise source using a predetermined learning model (first learning model).

[0070] More specifically, the first specific sound is stored in the storage unit 4 in association (labeled) with each operation of the noise source.

[0071] The second specific sound, which is a specific sound, is associated (labeled) with the worker's action using a predetermined learning model (first learning model) and stored in the storage unit 4.

[0072] More specifically, the second specific sound is individually associated (labeled) with each action of the worker and stored in the storage unit 4. Note that the second specific sound may be further associated with at least one of an object (e.g., a tool or a part) held by the worker or the posture of the worker and stored in the storage unit 4.

[0073] The personal information is, for example, information that enables authentication of a registered person, and is information that enables confirmation of the appearance of, for example, image information.

[0074] The determination unit 5 detects the movement of the noise source from the information obtained by capturing an image with the imaging unit 2, and determines whether the detected movement of the noise source is a movement stored in association with the first specific sound.

[0075] In more detail, the judgment unit 5 predicts the future behavior of the noise source from the detected behavior of the noise source, and judges whether the predicted future behavior of the noise source is a behavior stored in association with the first specific sound.

[0076] Furthermore, the determination unit 5 detects the worker's movement from the information obtained by capturing an image with the imaging unit 2, and determines whether the detected worker's movement is a movement stored in association with the second specific sound.

[0077] In more detail, the judgment unit 5 predicts the worker's future actions from the detected worker's actions, and judges whether the predicted worker's future actions are actions stored in association with the second specific sound.

[0078] These determinations made by the determination unit 5 are made using a predetermined learning model (second learning model) that has been learned in advance.

[0079] When it is determined that the detected movement of the noise source is a movement that is associated with a first specific sound and stored in advance, the noise prediction unit 6 generates noise that is predicted to be emitted from the noise source based on the associated first specific sound.

[0080] In more detail, the noise prediction unit 6 generates noise that is expected to be emitted from the noise source based on a first specific sound associated with the future behavior of the noise source predicted from the detected behavior of the noise source.

[0081] Furthermore, when the noise prediction unit 6 determines that the detected movement of the worker is a movement that is associated with a second specific sound and stored in advance, it generates a noise that is predicted to be emitted by the worker based on the associated second specific sound.

[0082] In more detail, the noise prediction unit 6 generates noise that is expected to be emitted from the worker based on a second specific sound associated with the worker's future actions predicted from the detected worker's actions.

[0083] The anti-noise generation unit 7 generates noise cancellation noise as a noise reduction sound that reduces the noise generated by the noise prediction unit 6. That is, the anti-noise generation unit 7 generates noise cancellation noise based on a predetermined first specific sound associated with the detected movement of the noise source and a predetermined second specific sound associated with the detected movement of the worker. In other words, the anti-noise generation unit 7 generates noise cancellation noise in accordance with the determination result of the determination unit 5.

[0084] The person detection unit 32 determines whether a pre-registered person is within a predetermined distance from the worker from the information obtained by capturing an image with the imaging unit 2. A pre-registered person is, for example, a person who may have a conversation with the worker, such as giving instructions to the worker.

[0085] When a registered person is within a predetermined distance from the worker, the noise processing unit 33 takes in an external sound and adds (superimposes) it onto noise for noise cancellation.

[0086] The anti-noise generating unit 7 is capable of adding external audio to noise for noise cancellation and transmitting the result to the output unit 8.

[0087] When the audio output device 31 of the third embodiment detects that a registered person has approached the worker, it captures external audio and outputs it to the worker together with noise for noise cancellation.

[0088] The voice output device 31 of the third embodiment can achieve substantially the same effects as the voice output device 1 of the first embodiment described above, and can also make it easier for the worker to hear the voice of a registered person speaking to the worker. In other words, the voice output device 31 can make it easier for the worker to hear instructions given by people around him.

[0089] FIG. 6 is a flowchart showing an example of the control flow in the audio output device 31 of the third embodiment.

[0090] In step S41, the imaging unit 2 captures an image of a noise source and the like, and the sound collection unit 3 collects (detects) sounds around the worker.

[0091] In step S42, the future movement of the noise source is predicted from the imaging result from the imaging unit 2. That is, in step S42, the future movement of the noise source and the worker is predicted.

[0092] In step S43, the specific sounds associated with the predicted noise source motion are referenced. That is, in step S42, the first specific sound associated with the predicted noise source motion and the second specific sound associated with the predicted worker motion are referenced.

[0093] In step S44, the noise predicted to be generated in the future from the noise source is predicted using as input the referenced specific sound and the sound collected by the sound collection unit 3. That is, in step S44, the noise predicted to be generated in the future from the noise source and the worker is predicted using as input the referenced first and second specific sounds and the sound collected by the sound collection unit 3.

[0094] In step S45, noise for noise cancellation is generated based on the noise predicted in step S44.

[0095] In step S46, it is determined whether or not a person who has been pre-registered has been detected from the imaging result from the imaging unit 2.

[0096] In step S47, the distance between the detected pre-registered person and the worker is determined.

[0097] In step S48, if the distance between the detected pre-registered person and the worker is equal to or less than a predetermined value, external sounds are taken in.

[0098] In step S49, if the distance between the pre-registered person and the worker is equal to or less than a predetermined value, the noise cancellation noise and the captured external sound are played back through the output unit 8. Note that if the distance between the pre-registered person and the worker is not equal to or less than the predetermined value, the noise cancellation noise is played back through the output unit 8 in step S49.

[0099] Although specific embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the invention.

[0100] For example, the second and third embodiments described above may be combined appropriately to remove noise in the high frequency range with high precision.

[0101] For example, if the noise source movement stored in association with a specific sound is part of a series of movements (repeated movements), it is possible to determine whether the detected noise source movement is part of a series of movements based on the data stored in memory unit 4 in association with the noise source movement and the specific sound, and if the detected noise source movement is part of a series of movements, to predict the noise source's subsequent movement and output noise for noise cancellation.

[0102] That is, if the motion of the noise source stored in association with the first specific sound is part of a series of motions (repetitive motions), it is possible to determine whether the detected motion of the noise source is part of the series of motions based on data stored in memory unit 4 that associates the motion of the noise source with the first specific sound, and if the detected motion of the noise source is part of the series of motions, it is possible to predict the subsequent motion of the noise source and output noise for noise cancellation. Also, if the motion of the worker stored in association with the second specific sound is part of a series of motions (repetitive motions), it is possible to determine whether the detected motion of the worker is part of the series of motions based on data stored in memory unit 4 that associates the motion of the worker with the second specific sound, and if the detected motion of the worker is part of the series of motions, it is possible to predict the subsequent motion of the worker and output noise for noise cancellation.

[0103] The above-described embodiments relate to an audio output method, an audio output device, and a control program for executing the audio output method. [Explanation of symbols]

[0104] 1...Audio output device 2...Image capture unit 3…Sound collection section 4...Storage section 5…Judgment section 6...Noise prediction section 7...Anti-noise generation section 8...Output section 9...Server

Claims

1. memorizing specific sounds associated with the operation of noise sources around the worker; an imaging unit attached to the worker captures an image of the noise source; Detects sounds around the worker, detecting the operation of the noise source from the information obtained by imaging by the imaging unit; determining whether the detected motion of the noise source is a motion stored in association with the specific sound; If it is determined that the detected motion of the noise source is a stored motion, a noise reduction sound is generated based on the specific sound; a sound output method for outputting the noise reduction sound to the worker;

2. 2. The voice output method according to claim 1, wherein the noise source includes a noise generating source around the worker.

3. 3. The audio output method according to claim 2, wherein the noise source includes the worker.

4. 2. The voice output method according to claim 1, wherein the imaging unit is a camera attached to the head of the worker.

5. The audio output method described in claim 3, wherein the first specific sound associated with the movement of the noise source and the second specific sound associated with the movement of the worker are stored in association with the movement of the noise source or the movement of the worker using a predetermined learning model.

6. The voice output method according to claim 3, characterized in that a predetermined learning model is used to determine whether the detected movement of the noise source and the movement of the worker are movements stored in association with the specific sound.

7. The voice output method according to claim 3, characterized in that the second specific sound, which is the specific sound associated with the worker's action, is stored in association with at least one of an object held by the worker or the worker's posture.

8. 2. The audio output method according to claim 1, wherein, when the motion of the noise source stored in association with the specific sound is part of a series of motions, it is determined whether the detected motion of the noise source is part of a series of motions based on data stored in association with the motion of the noise source and the specific sound, and when the detected motion of the noise source is part of a series of motions, a subsequent motion of the noise source is predicted and the noise reduction audio is output.

9. 2. The audio output method according to claim 1, wherein the specific sounds are a high-frequency specific sound and a low-frequency specific sound.

10. 2. The voice output method according to claim 1, wherein, when it is detected that a registered person is approaching the worker, an external voice is captured and output to the worker together with the noise reduction voice.

11. a memory unit that stores specific sounds associated with the actions of noise sources around the worker; an imaging unit worn by the worker to image the noise source; a sound collection unit that detects sounds around the worker; a detection unit that detects the motion of the noise source and the motion of the worker from information obtained by imaging by the imaging unit; a determination unit that determines whether the detected motion of the noise source is a motion stored in association with the specific sound; an anti-noise generating unit that generates a noise reduction sound based on the specific sound when the detected motion of the noise source is determined to be a stored motion; an output unit that outputs the noise reduction sound to the worker.

12. 2. A control program for audio output that causes a predetermined computer to execute the audio output method according to claim 1.

Images