Voice processing device, voice processing program, voice output device, and animal behavior control method

Abstract

To provide a method for controlling a control target animal by using recorded animal vocalizations, ensuring that the target animal can be controlled independently of individual variation.SOLUTION: A voice processing device (1) includes: an acquisition unit (141) that acquires data on vocalizations emitted by an animal when it is in a specific situation; an extraction unit (142) that extracts a specific section from the acquired vocalization data; a generation unit (143) that generates pseudo-vocalization data mimicking an animal's vocalization, based on the extracted specific section; and an output unit (12) that outputs the generated data or a pseudo-vocalization based on the data.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a sound processing device, a sound processing program, and a method for controlling the behavior of an animal, and more particularly to a sound processing device that synthesizes sound, a sound processing program for synthesizing sound, and a method for controlling the behavior of an animal by using the synthesized sound. [Background technology]

[0002] Various technologies that utilize the cries of animals have been proposed in the past. For example, Patent Document 1 describes a bird and animal damage prevention device that emits warning sounds toward birds and animals based on warning sounds recorded for each individual bird and animal. In addition, Non-Patent Document 1 describes a method of extracting acoustic information (fundamental frequency, power, Mel frequency cepstral coefficient, etc.) from the cries of cattle, classifying the cattle by age based on the acoustic information, and further identifying the individual cattle classified by age. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2019-154442 A [Non-patent literature]

[0004] [Non-Patent Document 1] Individual Cow Identification Considering Age Changes in Mooing, 34th National Conference of the Japanese Society for Artificial Intelligence, 2020 Summary of the Invention [Problem to be solved by the invention]

[0005] The sounds emitted by the conventional device described in Patent Document 1 were simply recordings, so there was a possibility that some individuals would not respond. Also, as described in Non-Patent Document 1, the conventional use of cow sounds was limited to individual identification.

[0006] One aspect of the present invention aims to provide a method for controlling an animal using collected animal cries, which enables the control of the animal regardless of whether the animal is an individual. [Means for solving the problem]

[0007] In order to solve the above problems, a voice processing device according to one embodiment of the present invention includes an acquisition unit that acquires data on cries made by a sound-making animal in a specific situation, an extraction unit that extracts a specific part from the acquired cries data, a generation unit that generates pseudo-cries data that resemble the cries of the animal based on the extracted specific part, and an output unit that outputs the generated data or a pseudo-cries based on the data.

[0008] A sound processing program according to another aspect of the present invention causes a processor to execute an acquisition step of acquiring data on cries made by a sound-making animal in a specific situation, an extraction step of extracting a specific portion from the acquired sound data, a generation step of generating data on a pseudo-sound that resembles the sound of the animal based on the extracted specific portion, and an output step of outputting the generated data or a pseudo-sound based on the data.

[0009] Another invention relates to an audio output device that includes a memory unit that stores data of simulated cries and a speaker that outputs simulated cries based on the data, the data being generated based on specific portions extracted from data of a plurality of cries made by a sound-making animal under specific circumstances.

[0010] A method for controlling animal behavior according to another aspect of the present invention includes an acquisition step of acquiring data on cries made by a vocal animal in a specific situation; an extraction step of using a processor to extract a specific portion from the acquired cries data; a generation step of using a processor to generate artificial cries data that resemble the cries of the animal based on the specific portion extracted; and an output step of outputting the generated data, or an artificial cries based on the data, and playing the generated data or the artificial cries based on the data for the animal to be controlled. Effect of the Invention

[0011] According to one aspect of the present invention, it is possible to control the animal to be controlled, regardless of the individual animal. [Brief description of the drawings]

[0012] [Figure 1] 1 is a block diagram showing an audio processing device according to an embodiment of the present invention; [Diagram 2] FIG. 2 is a diagram showing an example of a waveform of an animal's cry processed by the device. [Diagram 3] FIG. 13 is a block diagram showing an audio output device according to an embodiment of the present invention. [Figure 4] 13 is a flowchart showing the flow of an animal control method according to an embodiment of another invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] <Audio processing device> Hereinafter, a voice processing device according to an embodiment of the present invention will be described in detail.

[0014] [Configuration of the audio processing device] The voice processing device 1 generates a pseudo-call from a call made by an animal A1 (a target for collecting calls) that makes a sound, to be played by an animal A2 that is a target for control and has a relationship with the animal A1. The "animals that make a sound" include calves, bulls, deer, wild boars, etc. The "relationship" includes a parent-child relationship, a relationship between opposite sexes, a relationship belonging to the same herd, etc. The "control" includes calling an animal A2 that is far away to a nearby place, chasing an animal A2 away, etc. The "animals to be controlled" include livestock that is an animal that is desired to be called to a nearby place, and harmful animals that are an animal that is desired to be chased away away. The livestock include mother cows (cows), etc. The harmful animals include deer, wild boars, etc. As shown in FIG. 1, the voice processing device 1 includes an input unit 11, an output unit 12, a processing-side memory unit 13, and a processing-side control unit 14.

[0015] [Input section] Data on the animal's cry is input to the input unit 11 from a microphone M placed near the animal A1 making the sound. The input unit 11 is composed of, for example, a terminal to which a cable leading to the microphone M is connected, or a communication unit that communicates with the microphone M in a wired or wireless manner. The input unit 11 may be a drive that reads a medium on which the data on the animal's cry is stored, or may be the microphone itself.

[0016] [Output section] The output unit 12 according to this embodiment outputs the generated data of the simulated cry. The output unit 12 is composed of, for example, a terminal to which a cable connected to the audio output device 2 is connected, a communication unit that communicates with the audio output device 2 by wire or wirelessly, a drive that writes the cry data to a medium, and the like. The output unit 12 may be configured to output the simulated cry based on the data. In other words, the output unit 12 may be a speaker. In this way, it is not necessary to separately prepare a dedicated device (such as the audio output device 2 described below) for outputting the simulated cry.

[0017] [Processing side storage unit] The processing side storage unit 13 stores a voice processing program 131 in which the operation of the voice processing device 1 is described. The voice processing program 131 may be created from scratch, or may be created using existing software (e.g., Praat, etc.). The processing side storage unit 13 according to this embodiment is composed of a semiconductor memory, a hard disk, etc. The processing side storage unit 13 may be configured to be capable of storing at least one of the cry data and the simulated cry data input to the input unit 11. The program is

[0018] [Processing side control section] The processing-side control unit 14 includes an acquisition unit 141, an extraction unit 142, a generation unit 143, and an output processing unit 144. The processing-side control unit 14 according to this embodiment is configured with a processor. When this processor executes the voice processing program 131 stored in the processing-side storage unit 13, the voice processing program 131 causes the processing-side control unit 14 to function as each of the control blocks 141 to 144.

[0019] (Acquisition Department) The acquisition unit 141 executes an acquisition process. In the acquisition process, the acquisition unit 141 acquires data of cries that are respectively made by a plurality of animals A1 that emit sounds when they are in a specific situation. The "specific situation" differs depending on the type of the animal A1 that emits the sound. If the animal A1 that emits the sound is a calf, the specific situation is when it is weaned. "Weaning" refers to the time when a calf stops natural nursing and leaves its mother. If the animal A1 that emits the sound is a bull, the specific situation is when it is placed in a herd of female cows. If the animal A1 that emits the sound is a pest, the specific situation is when it senses a physical or psychological threat. That is, if the animal A1 that emits the sound is a pest, the acquisition unit 141 acquires data of an alarm sound. This alarm sound is a sound that can be made regardless of age or sex.

[0020] (Extraction part) The extraction unit 142 executes an extraction process. In the extraction process, the extraction unit 142 extracts specific parts from the acquired data of a plurality of cries. The cries are a combination of closed and open sounds. When the cries are visualized, a waveform such as that shown in the upper part of FIG. 2 is shown. A waveform with a relatively small amplitude from the beginning of the sound to the dashed line L indicates a closed sound, and a waveform with a relatively large amplitude after the dashed line L1 indicates an open sound. The extraction unit 142 extracts the open sound part (the part corresponding to the waveform on the right side of the dashed line L1) from the cries. Since the open sound is considered to be a biologically significant part of the cries, by extracting this part, it becomes possible to generate a pseudo cry that can control the animal A2 to be controlled with a higher probability.

[0021] The extraction unit 142 may be configured to extract a part of the open-mouth sound. When the waveform of the cry shown in the upper part of FIG. 2 is Fourier-transformed, a spectrogram (cry data in the frequency domain with the vertical axis representing frequency and the horizontal axis representing time) as shown in the lower part of FIG. 2 is obtained. The extraction unit 142 extracts the part of the cry represented by the spectrogram that is a straight line extending horizontally between the dashed line L2 and the dashed line L3. This straight line part is the part of the open-mouth sound that is clearly pronounced. Therefore, by doing this, the part of the open-mouth sound that is ambiguous in pronunciation is removed, so that it becomes possible to generate data of a pseudo cry with clearer pronunciation. From the spectrogram, the fundamental frequency, sound intensity, length of the voice, etc. can be easily quantified using commercially available sound analysis software. These numerical values ​​can be used to identify individuals, as well as to determine breed differences, sex differences, age differences, and language differences.

[0022] (Generation part) The generating unit 143 executes a generating process. In the generating process, the generating unit 143 generates data of a pseudo cry that resembles the cry of the animal A1 based on the extracted parts. The generating unit 143 according to the present embodiment generates an average waveform by averaging the waveforms of the parts extracted from each cry data. The generating unit 143 then generates data of the pseudo cry by matching a closing sound to the average waveform. In this way, the pseudo cry data can be generated more easily than when other calculation methods are used. The pseudo cry data generated from the cry of a calf or a bull becomes data for cows to attract mother (female) cows nearby. The pseudo cry data generated from the cry of a deer (wild boar) becomes data for deer (wild boar) to drive the deer (wild boar) away (pseudo alarm sound data).

[0023] Note that the generating unit 143 may be configured to generate data of artificial cries using pitch-synchronized overlapping addition, rather than generating an average waveform. In this case, the generating unit 143 first generates data of a plurality of synthetic sounds, each differing only in fundamental frequency, based on data of a certain cry. Specifically, the generating unit 143 changes only the fundamental frequency by a small amount (for example, about 10 Hz) from the initial fundamental frequency without changing the spectrogram (the shape of the voice including the closed sound and the open sound), and generates data of the synthetic sound. This synthetic sound is created repeatedly while changing the change amount of the frequency, and data of a plurality of synthetic sounds is generated. Then, the generating unit 143 generates an added waveform by overlapping and adding the data of the plurality of synthetic sounds. The generating unit 143 then generates the data of the artificial cries by matching the closed sound to the added waveform.

[0024] The generating unit 143 may be configured to generate only two or less types of data among data for cattle, deer, and wild boar, or may be configured to be capable of generating four or more types of data including data for animals other than cattle, deer, and wild boar. The generating unit 143 may be configured to separately generate data for cattle that is generated from the cry of a calf and that is generated from the cry of a bull.

[0025] (Output processing section) The output processing unit 144 executes an output process. In the output process, the output processing unit 144 controls the output unit 12 so that the output unit 12 outputs data of the artificial cry.

[0026] [Variations of the voice processing device] The sound processing device 1 may be configured so that the acquisition unit 141 acquires data of the extracted open-mouth sound. In this case, the processing-side control unit 14 does not need to include the extraction unit 142.

[0027] Also, some or all of the functions of the control blocks 141 to 144 can be realized by a logic circuit. For example, the scope of the present invention includes an integrated circuit in which a logic circuit that functions as each of the control blocks 141 to 144 is formed. In addition, the functions of the control blocks 141 to 144 can be realized by, for example, a quantum computer.

[0028] Furthermore, each process executed by each of the control blocks 141 to 144 may be executed by an AI (Artificial Intelligence). In this case, the AI ​​may be one that operates in the processing side control unit 14, or may be one that operates in another device (such as an edge computer or a cloud server) different from the voice processing device.

[0029] [Effects of the audio processing device] The sound processing device 1 described above generates data on simulated cries using data on cries made by multiple animals A1 that make sounds under specific circumstances. The simulated cries based on such data reflect the characteristics of each of the cries of the multiple animals A1. When the multiple animals A2 to be controlled hear the simulated cries, they mistake the simulated cries for cries made by individuals with whom they have a relationship, and they behave in accordance with the simulated cries. Therefore, according to the sound processing device 1, the simulated cries based on the generated data can be directly output in the location where the animal A2 to be controlled is present, or the data can be sent to a playback device to be output, thereby making it possible to control the animal A2 to be controlled regardless of the individual.

[0030] <Audio output device> Next, a sound output device 2 according to another embodiment of the present invention will be described.

[0031] [Audio output device configuration] The sound output device 2 is for emitting a simulated cry. As shown in Fig. 3, the sound output device 2 includes a speaker 21, an output side storage unit 22 (storage unit), an operation unit 23, and an output side control unit 24.

[0032] [Output side storage unit] The output side storage unit 22 stores the simulated cry data 221. The output side storage unit 22 according to this embodiment is composed of a semiconductor memory, a hard disk, etc. The data 221 stored in the output side storage unit 22 and storing the simulated cry data 221 is data generated by the above-mentioned sound processing device 1. In other words, the data 221 is generated based on a plurality of specific parts respectively extracted from data of a plurality of cries uttered by a plurality of individual animals that make sounds in a specific situation.

[0033] Moreover, the output side storage unit 22 according to this embodiment stores three types of simulated sound data 221 for cattle, deer, and wild boar. The simulated sound data 221 stored in the output side storage unit 22 may be only two types or less of data for cattle, deer, and wild boar, or may be four or more types of data including data for animals other than cattle, deer, and wild boar. Furthermore, the output side storage unit 22 may store, as cattle data, data generated from the sound of a calf and data generated from the sound of a bull separately.

[0034] [Operation unit] The operation unit 23 is operated by a user of the audio output device 2. The operation unit 23 according to this embodiment has three buttons. The buttons correspond to a cow, a deer, and a wild boar, respectively. The number of buttons provided in the operation unit 23 may be two or less, or may be four or more, depending on the type of simulated sound data stored in the output-side memory unit 22. The operation unit 23 may be configured with a switch, a sensor, or a touch panel.

[0035] [Output side control section] The output side control unit 24 calls up data of a simulated sound from the output side storage unit 22 based on an operation performed on the operation unit 23, and outputs the data to the speaker 21. As a result, the speaker 21 outputs a simulated sound based on the data. The output side control unit 24 according to this embodiment calls up data 221 of a simulated sound corresponding to the operated button from among multiple types of data 221 of simulated sounds stored in the output side storage unit 22, and provides the data 221 to the speaker 21. That is, when the button corresponding to a cow is operated, the output side control unit 24 provides the speaker 21 with data 221 of a simulated sound for a cow. Also, when the button corresponding to a deer (wild boar) is operated, the output side control unit 24 provides the speaker 21 with data 221 of a simulated sound for a deer (wild boar).

[0036] [Modifications of audio output device] The voice output device 2 may include an input unit to which the data of the imitation cry is input from the voice processing device 1. In this way, the data of the imitation cry can be easily updated simply by reading new data of the imitation cry (without replacing the output side storage unit 22). The audio output device 2 may also include a second operation unit and an amplifier. The second operation unit is for adjusting the volume operated by the user. The amplifier increases or decreases the volume of the simulated cries emitted by the speaker 21 in response to an operation performed on the second operation unit. In this way, the volume of the simulated cries output by the speaker 21 can be adjusted in accordance with the size of the place where the audio output device 2 is used and the number of animals A2 to be controlled.

[0037] [Action and effect of audio output device] The sound output device 2 can call the controlled animal A2 close or drive it away by making the animal A2 (the animal whose sound is collected) emit a sound that is a control target and has a relationship with the animal A1. Therefore, the sound output device 2 can control the controlled animal A2 regardless of the individual animal by outputting a simulated sound in the place where the controlled animal A2 is present. In particular, by making a simulated sound based on data for cows play to a large number of mother cows (cows) scattered far away in the pasture, the large number of mother cows (cows) can be easily gathered close. Also, by making a simulated sound for deer (wild boar) play to deer (wild boar) staying in the pasture, the deer (wild boar) can be easily driven away.

[0038] <Animal control methods> Next, an animal control method according to another embodiment of the present invention will be described.

[0039] [Flow of Animal Control Methods] As shown in FIG. 4, the animal control method includes a collection step S1, an extraction step S2, a generation step S3, and an output step S4.

[0040] (Collection step) In the first collection step S1, cries made by a plurality of animals A1 that make sounds in a specific situation are collected. The cries are collected using the above-mentioned sound processing device 1. Note that in the collection step S1, the cries may be collected using a general recording device, and the data of the cries may be transferred from the recording device to the sound processing device 1.

[0041] (Extraction step) After the calls are collected, the process proceeds to extraction step S2. In extraction step S2, specific parts are extracted from each of the collected call data. The extraction of the specific parts is performed using the above-mentioned voice processing device 1. Note that in the extraction step, the extraction of the specific parts may be performed using a general voice processing device different from the voice processing device 1.

[0042] (Generation step) After extracting the specific portion from the cry data, the process proceeds to a generation step S3. In the generation step S3, data of a pseudo cry that resembles the cry of the animal A1 is generated based on the extracted portions. The generation of the pseudo cry data is performed using the above-mentioned sound processing device 1.

[0043] (Output step) After the data of the simulated cry is generated, the process proceeds to an output step S4. In the output step S4, the simulated cry based on the generated data is played for the controlled animal A2. The simulated cry is played using the audio output device 2. Note that in the output step S4, the simulated cry may be played using a general playback device that stores the data of the simulated cry or that reads media that stores the data of the simulated cry.

[0044] [Effects of animal control methods] The above-described animal control method controls the target animal A2 by making the target animal A2 hear an artificial cry based on the data generated by the sound processing device 1. Therefore, according to this animal control method, the target animal A2 can be controlled regardless of the individual animal.

[0045] <Summary> The sound processing device according to aspect 1 of the present invention comprises an acquisition unit that acquires data on cries made by a sound-making animal in a specific situation, an extraction unit that extracts a specific part from the acquired sound data, a generation unit that generates artificial sound data that resembles the animal's sound based on the extracted specific part, and an output unit that outputs the generated data or an artificial sound based on the data.

[0046] The voice processing device according to aspect 2 of the present invention may be configured in the above aspect 1 such that the animal making the sound is a calf, and the specific situation is when the calf is weaned.

[0047] A voice processing device according to aspect 3 of the present invention may be configured in the above-mentioned aspect 1 such that the animal making the sound is a bull, and the specific situation is being placed in the midst of a herd of female cows.

[0048] The sound processing device according to aspect 4 of the present invention may be configured in any one of aspects 1 to 3 above, so that the acquisition unit acquires data on the cries made by a plurality of individual animals that make sounds in a specific situation, the extraction unit extracts the specific parts from the acquired data on the plurality of cries, and the generation unit generates data on the simulated cries based on the extracted specific parts.

[0049] A sound processing device according to aspect 5 of the present invention may be configured in any one of aspects 1 to 4 above, such that the cry is a combination of a closed mouth sound and an open mouth sound, and the extraction unit extracts the open mouth sound portion of the cry as the specific portion.

[0050] A sound processing device according to a sixth aspect of the present invention may be configured in the above-mentioned fifth aspect such that the generation unit generates an average waveform by averaging the waveforms of the extracted multiple parts, and generates the artificial cry data by matching the closing sound to the average waveform.

[0051] A sound processing device according to aspect 7 of the present invention may be configured in any one of aspects 1 to 3 above, such that the cry is a combination of a closed sound and an open sound, the extraction unit extracts the open sound portion of the cry as the specific portion, and the generation unit generates data for a plurality of synthetic sounds, each differing only in fundamental frequency, based on the specific portion extracted from data for a certain cry, generates an additive waveform by superimposing and adding the data for the plurality of synthetic sounds, and generates the artificial cry data by combining the closed sound with the additive waveform.

[0052] The audio processing program according to aspect 8 of the present invention is configured to cause a processor to execute an acquisition step of acquiring data on cries made by a sound-making animal in a specific situation, an extraction step of extracting a specific portion from the acquired sound data, a generation step of generating data on an imitation sound that resembles the sound of the animal based on the extracted specific portion, and an output step of outputting the generated data or an imitation sound based on the data.

[0053] The audio output device of aspect 9 of the present invention comprises a memory unit that stores data of simulated cries, and a speaker that outputs simulated cries based on the data, wherein the data is generated based on specific parts extracted from data of a plurality of cries made by a sound-making animal in a specific situation.

[0054] The animal behavior control method according to aspect 10 of the present invention is a method including an acquisition step of acquiring data on cries made by a vocal animal in a specific situation; an extraction step of using a processor to extract a specific portion from the acquired cries data; a generation step of using a processor to generate artificial cries data that resemble the cries of the animal based on the specific portion extracted; and an output step of outputting the generated data, or an artificial cries based on the generated data, and playing the generated data to the animal to be controlled.

[0055] The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of symbols]

[0056] 1 Audio processing device 11 Input section 12 Output section 13 Processing side memory unit 131 Audio Processing Programs 14 Processing side control section 141 Acquisition Department 142 Extraction part 143 Generation part 144 Output Processing Section 2 Audio output device 21 Speaker 22 Output side memory unit 221 Artificial cry data 23 Control section 24 Output side control section A1 Animals (sounds collected) A2 Animals (Controlled) M Microphone

Claims

1. An acquisition unit that acquires data on cries made by a vocal animal in a specific situation; An extraction unit that extracts specific parts from the acquired bird call data; a generator that generates data of a pseudo-sound that resembles the sound of the animal based on the extracted specific portion; an output unit that outputs the generated data or a simulated cry based on the data; Equipped with 13. A voice processing device comprising:

2. the vocalizing animal is a calf; The particular situation is weaning.

2. The audio processing device according to claim 1 .

3. the vocalizing animal is a bull; The particular situation is being placed in a herd of cows.

2. The audio processing device according to claim 1 .

4. The acquisition unit acquires data on cries made by each of a plurality of animals that make sounds in a specific situation, The extraction unit extracts the specific parts from the acquired data of the plurality of cries. The generation unit generates the artificial cry data based on the extracted specific parts.

4. The audio processing device according to claim 1, wherein the audio processing device is a digital audio processing device.

5. The cry is a combination of a closed sound and an open sound, The extraction unit extracts the open-mouth sound portion from the cry as the specific portion.

5. The audio processing device according to claim 4.

6. The generation unit is generating an average waveform by averaging the waveforms of the plurality of extracted portions; generating the artificial cry data by combining the closing sound with the average waveform; 6. The audio processing device according to claim 5,

7. The cry is a combination of a closed sound and an open sound, The extraction unit extracts the open-mouth sound portion from the cry as the specific portion, The generation unit is generating a plurality of synthetic sound data each differing only in fundamental frequency based on the specific portion extracted from a certain bird cry data; generating an added waveform by superimposing and adding the data of the plurality of synthetic sounds; The artificial bark data is generated by combining the closing sound with the added waveform.

4. The audio processing device according to claim 1, wherein the audio processing device is a digital audio processing device.

8. The processor: An acquisition step of acquiring data of a call made by a vocal animal in a particular situation; An extraction step of extracting a specific portion from the acquired call data; A generating step of generating data of a pseudo-sound that resembles the sound of the animal based on the extracted specific portion; an output step of outputting the generated data or a simulated cry based on the data; Execute the A speech processing program comprising:

9. A storage unit for storing data of the simulated cry; a speaker that outputs a simulated cry based on the data; Equipped with The data is generated based on a specific portion extracted from data of a plurality of calls made by a vocalizing animal in a specific situation.

13. An audio output device comprising:

10. An acquisition step of acquiring data of a call made by a vocal animal in a particular situation; An extraction step of extracting a specific portion from the acquired call data using a processor; a generating step of generating data of an artificial sound that resembles the sound of the animal, using a processor, based on the extracted specific portion; an output step of outputting the generated data or an artificial cry based on the data and playing the generated data for the controlled animal; Including, A method for controlling animal behavior comprising:

Images