System and method for delivering voice, and computer readable non-transitory storage medium

Abstract

The invention relates to a system, a method and a computer-readable non-transitory storage medium for delivering voice. The present disclosure provides a technique for ensuring ease of discrimination between an artificial voice and a human voice in a system for delivering the artificial voice generated by a machine learning model to a user of a vehicle. The invention provides a voice transmission system. This voice transmission system is provided with one or more processors. One or more processors generate artificial speech that mimics the speech of the person through a machine learning model. The generated artificial speech is then delivered to a user of the vehicle. At this time, the processor changes a second parameter relating to the nature of the artificial voice in conjunction with a first parameter relating to at least one of the movement of the vehicle and the amount of operation by the driver of the vehicle.

Description

Technical Field

[0001] The present disclosure relates to a technology for delivering voice to a user of a vehicle, and more particularly, to a technology for delivering artificial voice generated by a machine learning model to a user of a vehicle. Background Art

[0002] Patent document 1 discloses a vehicle-mounted device that plays voice to a user riding in a vehicle based on voice information data for speech received from a roadside device. The vehicle-mounted device includes a speech speed determination unit and a voice signal generation unit. When playing voice based on the received voice information data for speech, the speech speed determination unit determines a speech speed corresponding to the situation. In addition, the voice signal generation unit generates a voice signal based on the voice information data for speech in a manner that achieves the determined speech speed. Then, a voice based on the generated voice signal is output to the user from a speaker or the like. According to the speech speed determination unit, in order to more reliably alert the user, when the voice information data for speech is data with a high degree of urgency, the speech speed is set high.

[0003] It should be noted that, as documents showing the technical level at the time of filing in the technical field of the present disclosure, in addition to the above-mentioned Patent Document 1, the following Patent Document 2 and Patent Document 3 can also be cited as examples.

[0004] Prior art literature

[0005] Patent Literature

[0006] Patent Document 1: Japanese Patent Application Publication No. 2007-279975

[0007] Patent Document 2: Japanese Patent Application Publication No. 2001-033256

[0008] Patent Document 3: Japanese Patent Application Publication No. 2004-046400

[0009] Consider applying a machine learning model to a device that transmits voice to a user of a vehicle, and outputting artificial voice generated by the machine learning model to the user. By applying the machine learning model, the artificial voice can be made close to the voice of a real person. On the other hand, it may be difficult for the user who hears the artificial voice to distinguish the artificial voice from the actual human voice. However, from the perspective of transparency of the system to which the machine learning model is applied, it is ideal that the distinction between the artificial voice generated by the machine learning model and the human voice is easy for the user. Summary of the invention

[0010] One object of the present disclosure is to provide a technology for ensuring ease of distinguishing artificial speech from human voice in a system for delivering artificial speech generated by a machine learning model to a user of a vehicle.

[0011] The present disclosure provides a speech transmission system. The speech transmission system of the present disclosure has one or more processors. The one or more processors generate artificial speech that imitates human speech through a machine learning model. Then, the generated artificial speech is transmitted to a user of a vehicle. At this time, the processor changes a second parameter related to the nature of the artificial speech in conjunction with a first parameter related to at least one of the movement of the vehicle and the operation amount of the driver of the vehicle.

[0012] In the voice transmission system of the present disclosure, the first parameter may be set to the acceleration of the vehicle, and the second parameter may be set to the height of the artificial voice. Furthermore, the processor may link the first parameter and the second parameter so that the greater the acceleration, the higher the height of the artificial voice.

[0013] The present disclosure provides a method for delivering a voice. The method for delivering a voice includes the following steps. The first step is to generate an artificial voice that imitates human voice through a machine learning model. The second step is to deliver the artificial voice to a user of a vehicle. The third step is to change a second parameter related to the nature of the artificial voice in conjunction with a first parameter associated with at least one of the movement of the vehicle and the operation amount of the driver of the vehicle.

[0014] The present disclosure provides a non-transitory storage medium storing a voice transmission program executed by a computer. The voice transmission program enables the computer to execute the above-mentioned voice transmission method.

[0015] Effects of the Invention

[0016] According to the technology disclosed in the present invention, the parameters related to the properties of artificial voice are changed in a manner linked to the parameters related to the movement of the vehicle or the operation amount of the driver of the vehicle. In this way, since the voice changes in a manner linked to the parameters related to the vehicle, it does not occur in the voice uttered by an actual person, it is possible to prevent the user of the vehicle from misidentifying the artificial voice emitted by the voice transmission system as the voice of an actual person. In addition, it is not necessary to set the artificial voice to a mechanical voice in order to prevent the user from misidentifying, so the user's sense of disharmony with the artificial voice can be suppressed.

[0017] In this way, according to the technology disclosed in the present invention, the sense of discomfort felt by the user can be suppressed, and the artificial voice generated by the machine learning model can be easily distinguished from the actual human voice. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1It is a block diagram showing a configuration example of a vehicle according to the present embodiment.

[0019] Figure 2 This is a conceptual diagram showing an example of a scenario in which speech transmission is performed by the speech transmission system according to the present embodiment.

[0020] Figure 3 is a block diagram representing an example of processing performed by a speech delivery system.

[0021] Figure 4 This is a table showing examples of combinations of the first parameter and the second parameter.

[0022] Figure 5 is a block diagram representing another example of processing performed by the speech delivery system.

[0023] Figure 6 is a block diagram showing yet another example of processing performed by the voice delivery system.

[0024] Description of Reference Numerals

[0025] 1: vehicle, 10: computer, 11: processor, 12: storage device, 13: program storage area, 14: model data storage area, 20: sensor group, 21: vehicle state sensor, 22: identification sensor, 23: position sensor, 30: speaker, 40: input device, 50: microphone, 100: voice transmission system, 200: automatic driving device. DETAILED DESCRIPTION

[0026] Embodiments of the present disclosure will be described with reference to the drawings.

[0027] 1. Overview of Voice Transmission System

[0028] The speech transmission system of this embodiment is a system that generates artificial speech using a machine learning model and transmits the generated artificial speech to a user of a vehicle. The user to whom the artificial speech is transmitted may be the driver of the vehicle or a passenger other than the driver. Figure 1 1 is a diagram showing a configuration example of a vehicle 1 to which the voice transmission system 100 according to the present embodiment is applied. The vehicle 1 may be a vehicle owned by an individual or a shared vehicle used for various mobility services. In addition, the vehicle 1 may also be a vehicle that is automatically driven by an automatic driving device.

[0029] The vehicle 1 includes a computer 10 for generating artificial speech, a sensor group 20, and a speaker 30. The computer 10 is connected to the sensor group 20 and the speaker 30 via an in-vehicle network. The speech transmission system 100 includes at least the computer 10. However, the speech transmission system 100 may include the speaker 30 in addition to the computer 10. In addition, the speech transmission system 100 may further include the sensor group 20.

[0030] The computer 10 is, for example, an ECU (Electronic Control Unit) mounted on the vehicle 1 or a collection of multiple ECUs. Alternatively, a part or all of the functions of the computer 10 may also be configured on an external server. In this case, the vehicle 1 is connected to the external server via a wireless communication network. In short, the computer 10 includes one or more processors 11 (hereinafter, simply referred to as processors 11) and one or more storage devices 12 (hereinafter, simply referred to as storage devices 12).

[0031] The processor 11 executes various processes. Examples of the processor 11 include a CPU (Central Processing Unit), an ASIC (Application Specific Integrated Circuit), and an FPGA (Field-Programmable Gate Array). The processor 11 may be a single processor or may include multiple processors.

[0032] As the storage device 12, HDD (Hard Disk Drive), SSD (Solid State Drive), volatile memory, non-volatile memory, etc. are exemplified. The storage device 12 may be a single storage device or may include multiple storage devices. The storage device 12 includes at least a program storage area 13 and a model data storage area 14. Each storage area may be implemented by a single storage device or by separate storage devices.

[0033] One or more programs are stored in the program storage area 13. Each program is composed of a plurality of instructions. Various processes implemented by the speech transmission system 100 are implemented by the cooperation between the processor 11 executing the program and the storage device 12. The programs stored in the program storage area 13 include at least a speech transmission program for causing the processor 11 to execute a process associated with the transmission of artificial speech. The speech transmission program can also be recorded in a computer-readable recording medium.

[0034] Model data is stored in the model data storage area 14. The model data stored in the model data storage area 14 is data of a machine learning model used for generating artificial speech.

[0035] The storage device 12 may include a storage area other than the program storage area 13 and the model data storage area 14. In the storage area other than the program storage area 13 and the model data storage area 14, for example, map information is stored.

[0036] The sensor group 20 includes a vehicle state sensor 21. The vehicle state sensor 21 detects the state of the vehicle 1. Specifically, the vehicle state sensor 21 detects parameters such as the speed, acceleration, steering angle, roll angular velocity, pitch angular velocity, yaw angular velocity, and tilt angle of the vehicle 1. As examples of the vehicle state sensor 21, a vehicle speed sensor, an acceleration sensor, a yaw angular velocity sensor, a steering angle sensor, a roll angular velocity sensor, a pitch angular velocity sensor, a yaw angular velocity sensor, and a tilt angle sensor are shown. In addition, the sensor group may also include an identification sensor 22 and a position sensor 23. The identification sensor 22 identifies the surrounding conditions of the vehicle 1. As examples of the identification sensor 22, a camera, LiDAR (Laser Imaging Detection and Ranging), and a radar are shown. As examples of the camera, a front camera that shoots the front of the vehicle 1, a rear camera that shoots the rear of the vehicle 1, and a side camera that shoots the side of the vehicle 1 are shown. The position sensor 23 detects the position and orientation of the vehicle 1. As the position sensor 23 , a GNSS (Global Navigation Satellite System) sensor is exemplified.

[0037] The computer 10 obtains the information detected by these sensors through the in-vehicle network. For example, the computer 10 obtains the state of the vehicle 1 from the vehicle state sensor 21. In addition, the computer 10 can also obtain the information detected by the identification sensor 22 and the position sensor 23. Moreover, the computer 10 can also perform processing such as sensing objects existing around the vehicle 1 and measuring the relative position and relative speed of the vehicle 1 relative to the sensed objects based on the information obtained from the identification sensor 22.

[0038] The speaker 30 is a speaker that transmits the artificial voice generated by the computer 10 to the user of the vehicle 1. The speaker 30 may include a plurality of speakers. The data of the artificial voice generated by the computer 10 is sent to the speaker 30 through the in-vehicle network. Then, the speaker 30 plays the voice based on the received data, thereby transmitting the artificial voice to the user.

[0039] The above describes an example of the configuration of the speech transmission system 100. Next, the transmission of artificial speech by the speech transmission system 100 will be described using a specific scenario as an example. Figure 2 This is a conceptual diagram showing an example of a scenario in which artificial voice is delivered.

[0040] exist Figure 2 In the example, the speaker 30 includes three speakers, namely, the speaker 30a, the speaker 30b, and the speaker 30c. And, the speaker 30a among the three speakers is used to transmit the artificial voice to the user. In this way, when the speaker 30 includes a plurality of speakers, the speaker that emits the voice at one time may be a single speaker. Alternatively, the voice may be emitted from a plurality of speakers at the same time.

[0041] In addition, Figure 2 In the example shown, vehicle 1 is an autonomous driving vehicle, and a voice message such as "Start autonomous driving" is transmitted in order to inform the user that the autonomous driving of vehicle 1 has started. In this way, information can be sent to the user of vehicle 1 through the transmission of artificial voice by the voice transmission system 100. Other examples of the content of the transmitted artificial voice include messages such as "Start moving", "Change lanes", "Arrive at destination soon", "Approaching a sharp turn", "There is an intersection ahead. Watch out for pedestrians rushing out", "Approaching the leading vehicle. Please keep a sufficient distance between vehicles", and "Entering a road with a different speed limit. Change speed". However, the transmitted artificial voice is not limited to a voice that provides information, and may also be a voice that provides certain attention reminders, such as a voice that is emitted to prompt the user to pay attention when the user's drowsiness is sensed.

[0042] In order to convey these messages, the computer 10 generates artificial voices that imitate human voices using a machine learning model. Here, with the advancement of technology in recent years, artificial voices generated using machine learning models can be close to real human voices. By making artificial voices close to real human voices, the sense of disharmony felt by users receiving the voices can be reduced compared to the case of playing mechanical voices. On the other hand, when users hear voices, they may find it difficult to distinguish whether it is an artificial voice or a real human voice. For example, in Figure 2 In the scenario, the driver of the vehicle 1 who hears the voice from the speaker 30a may mistakenly recognize the voice as the voice of another person riding in the vehicle 1, such as the voice of the passenger sitting in the front passenger seat. Alternatively, the voice may be mistakenly recognized as the voice transmitted from a person outside the vehicle 1 via wireless communication.

[0043] From the perspective of transparency of the system to which the machine learning model is applied, it is ideal that the artificial voice generated by the machine learning model is easily distinguishable from the actual human voice for the user of the vehicle 1. However, in order to reduce the sense of disharmony felt by the user who hears the artificial voice, the artificial voice itself is required to be not a mechanical voice but a voice close to a human voice.

[0044] People can judge whether the voice they hear is a real person's voice or an artificial voice by the naturalness of the change in the nature of the voice. The nature of the voice, such as volume, height, sound quality, etc., will also change in the daily conversations of real people, but the change does not follow certain regular changes. For example, factors that change the nature of a person's voice in a conversation can include the topic, the attitude of the other party in the conversation, changes in the surrounding environment, etc., but these factors are not factors that give regularity to the change in the nature of the sound. Therefore, when the change in the voice heard has certain regularity, people will instinctively feel that the change in the voice is unnatural, and thus can unconsciously judge that the voice is not the voice of an actual person, that is, the voice is an artificial voice.

[0045] Therefore, the voice transmission system 100 of the present embodiment acquires a parameter related to the movement of the vehicle as a first parameter. Then, a second parameter related to the property of the artificial voice is changed in a manner linked to the first parameter. By changing the second parameter regularly in linkage with the first parameter, the artificial voice transmitted by the voice transmission system 100 can be easily distinguished from the actual human voice.

[0046] 2. Processing example

[0047] Figure 3 11 is a block diagram showing an example of processing performed by the processor 11 by executing the speech transmission program. By executing the speech transmission program, the processor 11 performs processing 111, processing 112, processing 113, processing 114, and processing 115.

[0048] First, the processor 11 executes the process 111. The process 111 detects a scene in which the artificial voice is to be delivered. The scene in which the artificial voice is to be delivered is, for example, a scene in which the automatic driving of the vehicle 1 is started. In this case, the processor 11 determines the timing of the start of the automatic driving based on the information received from the automatic driving device 200 that performs the automatic driving of the vehicle 1. Then, when it is determined that it is the timing to start the automatic driving, the scene is detected as a scene in which the artificial voice is to be delivered.

[0049] The processor 11 may also detect a scene where artificial voice transmission is to be performed based on information received from a device other than the automatic driving device 200, such as the recognition sensor 22. For example, the processor 11 calculates the distance between the vehicle 1 and the preceding vehicle based on information obtained by recognizing the preceding vehicle by the recognition sensor 22. Then, when the inter-vehicle distance between the vehicle 1 and the preceding vehicle is less than a predetermined distance, it is detected that the scene is to be transmitted with artificial voice.

[0050] Alternatively, the processor 11 may also detect a scenario where artificial voice transmission is to be performed based on the map information stored in the storage device 12 and the position information detected by the position sensor 23. For example, the processor 11 obtains the current position of the vehicle 1 from the position sensor 23. In addition, the processor 11 obtains the destination of the vehicle 1 registered in the map information. Then, when it is determined based on this information that the vehicle 1 is approaching the destination, the scenario is detected as one where artificial voice transmission is to be performed. As another example, when the vehicle 1 is approaching an intersection or a sharp turn, the processor 11 may detect a scenario where artificial voice transmission is to be performed based on the map information and the position information of the vehicle 1.

[0051] It should be noted that the processor 11 may also use a machine learning model to perform processing 111. That is, the machine learning model may also be used to detect the scenario in which the artificial voice is to be transmitted.

[0052] Next, the processor 11 performs processing 112. The content of the utterance by the artificial voice is determined by processing 112. The content of the artificial voice is determined according to the scene detected by processing 111. The content of the artificial voice corresponding to each scene is pre-stored in the storage device 12, for example. Alternatively, the processor 11 may perform processing 112 using a machine learning model. That is, the content of the utterance by the artificial voice may also be determined by a machine learning model.

[0053] The processor 11 executes the processes 113 and 114 in parallel with the processes 111 and 112. The process 113 is used to obtain a parameter related to the motion of the vehicle 1 as a first parameter. Examples of the parameter related to the motion of the vehicle 1 include the speed, acceleration, steering angle, yaw rate, roll rate, pitch rate, and yaw rate of the vehicle 1. The processor 11 may obtain these parameters from the vehicle state sensor 21.

[0054] Next, the processor 11 executes the process 114. The second parameter is determined by the process 114. The second parameter is a parameter related to the property of the artificial speech generated by the speech transmission system 100. As an example of the second parameter, a parameter indicating the size (volume, sound pressure, amplitude), height (frequency), sound quality (waveform), etc. of the artificial speech is cited. The second parameter is determined in a manner linked to the first parameter obtained by the process 113.

[0055] Figure 4 An example of a combination of the first parameter and the second parameter is shown in the table. The processor 11, for example, links the acceleration of the vehicle 1 with the height of the artificial voice, and determines the height of the artificial voice in such a way that the greater the acceleration of the vehicle 1, the higher the voice of the artificial voice becomes. Or, for example, links the speed of the vehicle 1 with the volume of the artificial voice, and determines the volume of the artificial voice in such a way that the higher the speed of the vehicle 1, the larger the voice of the artificial voice becomes. Or, for example, links the yaw angular velocity of the vehicle 1 with the sound quality of the artificial voice, and changes the sound quality of the artificial voice as the yaw angular velocity of the vehicle 1 increases or decreases.

[0056] It should be noted that Figure 4 The combination shown in the table is an example, and any parameters can be selected and combined as the first parameter and the second parameter. In addition, the linked parameters do not have to be a one-to-one combination, and multiple second parameters can be linked to one first parameter. For example, the height and volume of the artificial voice can also be changed in a manner linked to the acceleration of the vehicle 1.

[0057] The combination of the first parameter and the second parameter and the corresponding relationship indicating how the two parameters are linked are stored in advance in the storage device 12. Alternatively, the processor 11 may also use a machine learning model to perform processing 114. That is, the second parameter may also be determined by the machine learning model in a manner linked to the first parameter.

[0058] Refer again Figure 3 . After the execution of processes 112 and 114, the processor 11 executes process 115. Through process 115, an artificial voice based on the content determined by process 112 and the parameters determined by process 114 is generated. In process 115, the processor 11 uses a machine learning model to generate an artificial voice that imitates human voice. It should be noted that the method of synthesizing an artificial voice that reads fixed lines using a machine learning model is well known, so it is omitted here. The data of the generated artificial voice is sent to the speaker 30, and the artificial voice is output from the speaker 30.

[0059] Through the above processing, the voice transmission implemented by the voice transmission system 100 is performed. As shown in the above processing, according to the voice transmission system 100 of this embodiment, when the voice is transmitted, the second parameter related to the property of the artificial voice is changed in a manner linked to the first parameter related to the movement of the vehicle 1. As a result, it is easy for the user to determine that the voice transmitted from the voice transmission system 100 is an artificial voice. That is, by generating such a change in the property of the voice that does not occur in the actual human voice, the user can easily perceive that the voice emitted by the voice transmission system 100 is an artificial voice, thereby preventing the user from misidentifying the voice as a voice emitted by a human.

[0060] Furthermore, according to the speech transmission system 100 of this embodiment, the second parameter is not simply changed, but the second parameter is changed in a manner linked with the first parameter. This makes it easier for the user to determine that the speech transmitted by the speech transmission system 100 is artificial speech.

[0061] The movement of the vehicle 1 that determines the first parameter is a movement that can be felt as a physical sensation for the user riding in the vehicle 1. Therefore, the user can easily perceive that the change in the second parameter is a change linked to the first parameter. In addition, it does not usually occur in human voice that the nature of the voice changes in a manner linked to the parameters related to the behavior of the vehicle 1. Therefore, it is easier for the user to know that the change in the second parameter is not a change caused by simple chance, but a change caused by the voice transmission system 100.

[0062] It should be noted that, as one of the methods for easily distinguishing the artificial voice generated by the system from the actual human voice, it is also considered to change the artificial voice into a mechanical voice, which is greatly different from the human voice. However, for the user receiving the voice, the voice close to the human voice is more likely to make him feel pleasant to the ear than the mechanical voice. In the voice transmission system 100 of the present embodiment, the ease of distinguishing the artificial voice can be improved while keeping the artificial voice as the voice close to the human voice, so while ensuring the transparency of the system, the user's satisfaction can also be improved.

[0063] 3. Modifications

[0064] 3-1. First Modification Example (Modification Example Related to the First Parameter)

[0065] A parameter other than the parameter related to the movement of the vehicle 1 may be used as the first parameter. For example, a parameter related to the operation amount of the driver of the vehicle 1 may be used as the first parameter. Figure 51 is a block diagram showing an example of processing based on a voice transmission program when a parameter related to an operation amount of a driver of a vehicle 1 is set as a first parameter. Here, in addition to being connected to a sensor group 20 and a speaker 30 through an in-vehicle network, the computer 10 is also connected to an input device 40 through the in-vehicle network. The input device 40 is a device for the driver of the vehicle 1 to input an operation amount. Examples of the operation amount input by the driver include an acceleration operation amount, a brake operation amount, a steering operation amount, and the like. Examples of the input device 40 include an accelerator pedal, a brake pedal, a steering wheel, and the like.

[0066] Processing 221 and Processing 222 and Figure 3 The processing 111 and the processing 112 are the same as those of FIG. 223 and the processing 224 are executed in parallel with the processing 221 and the processing 222, and the first parameter is obtained by the processing 223. Figure 3 However, the first parameter acquired by the process 223 is set as a parameter related to the operation amount of the driver of the vehicle 1. The processor 11 acquires the operation amount of the driver from the input device 40 and sets it as the first parameter, or calculates the first parameter based on the acquired operation amount.

[0067] The first parameter may be a parameter representing the operation amount of the driver. For example, a parameter representing the magnitude of the acceleration amount input to the input device 40 by the driver may be set as the first parameter.

[0068] Alternatively, the difference between the ideal operation amount and the operation amount actually input by the driver can also be set as the first parameter. The processor 11 can calculate the ideal operation amount based on the information obtained from the sensor group 20. For example, the operation amount of keeping the speed of the vehicle 1 constant can be set as the ideal operation amount. In this case, the processor 11 can calculate the ideal operation amount based on the acceleration of the vehicle 1 obtained from the vehicle state sensor 21. Or, for example, the operation amount of keeping the distance between the vehicle and the preceding vehicle constant can also be set as the ideal operation amount. The processor 11 calculates the relative distance between the vehicle 1 and the preceding vehicle based on the information obtained by the identification sensor 22 identifying the preceding vehicle. Then, the acceleration operation amount can be set as the ideal operation amount, and the accelerator operation amount is the operation amount of keeping the relative distance with the preceding vehicle constant. Alternatively, the processor 11 can also calculate the operation amount taking into account the change in the slope as the ideal operation amount based on the inclination angle of the vehicle 1 obtained from the vehicle state sensor 21, in such a way that the acceleration operation amount becomes larger when the vehicle 1 is traveling on an uphill road, and the braking operation amount becomes larger when the vehicle 1 is traveling on a downhill road.

[0069] The parameter associated with the operation amount of the driver is set as the first parameter, which means that the parameter associated with the operation input by the driver himself as the user of the vehicle 1 is set as the first parameter. That is, the first parameter is set as a parameter associated with the amount that the driver can grasp by himself. Therefore, the driver can more easily perceive that the second parameter associated with the artificial voice changes in a manner linked to the first parameter. In this way, it is easier for the user to determine that the voice transmitted by the voice transmission system 100 is an artificial voice obtained using a machine learning model, thereby improving the transparency of the system.

[0070] Furthermore, when the difference between the ideal operation amount and the operation amount actually input by the driver is set as the first parameter, the following effect is also expected. For the purpose of explanation, as an example of a specific scenario, consider a scenario in which the voice transmission system 100 transmits an artificial voice in order to remind the driver to pay attention to whether the acceleration operation amount is small or large. For example, there is a scenario in which when the vehicle 1 is approaching an uphill slope, the acceleration operation amount becomes insufficient without the driver noticing, and the vehicle 1 decelerates. It is assumed that in such a scenario, the driver is reminded to pay attention by artificial voice so that the driver steps on the accelerator.

[0071] Here, the first parameter is set to the difference between the ideal accelerator operation amount and the actually input accelerator operation amount, and the second parameter is set to the height of the artificial voice. The height of the artificial voice changes according to the difference between the driver's accelerator operation amount and the ideal accelerator operation amount. The voice transmission system 100 can not only convey the insufficient accelerator operation amount through the content conveyed by the artificial voice, but also convey the magnitude of the accelerator that the driver should step on through the change in the height of the artificial voice. That is, it is also possible to obtain an effect that the driver can easily and intuitively grasp the insufficient operation amount.

[0072] As another example of the first parameter, the distance between the vehicle 1 and an object existing in front of the vehicle 1 may be used. An example of the object existing in front of the vehicle 1 is a preceding vehicle.

[0073] However, it is ideal that the first parameter is set to a parameter that the user can feel. This is because, by setting the first parameter to a parameter that the user can easily feel the change through physical sensation, the user can easily perceive that the second parameter is linked to the first parameter, which is more effective in preventing the user from misidentifying the artificial voice as a human voice. Therefore, as the first parameter, a parameter related to the movement of the vehicle 1 and the operation amount of the driver of the vehicle 1 is optimal.

[0074] 3-2. Second Modification Example (Modification Example Related to the Second Parameter)

[0075] A parameter other than the parameter related to the properties of the artificial voice may be used as the second parameter. For example, the utterance speed of the artificial voice may be used as the second parameter. Assume that in this case, for example, the first parameter is set to the speed of the vehicle 1, and the first parameter and the second parameter are linked in such a way that the utterance speed becomes faster as the vehicle speed increases.

[0076] Alternatively, when the speaker 30 includes a plurality of speakers, the second parameter may be set to a parameter indicating the position of the speaker that plays the artificial voice. Figure 2 In this case, multiple speakers 30 are arranged on the left, right and center of the vehicle. In this case, the first parameter can be set to the steering angle of the vehicle 1, and the second parameter can be set to the position of the speaker that plays the artificial voice, so that the steering angle and the position of the speaker are linked in such a way that the sound source moves left and right as the steering angle changes left and right. Specifically, it is assumed that when the steering angle increases to the right, the artificial voice is emitted from the speaker 30a on the right, when the steering angle increases to the left, the artificial voice is emitted from the speaker 30c on the left, and when the steering angle approaches zero, the artificial voice is emitted from the speaker 30b in the center. In this way, it is easy for the user to know the change of the second parameter linked to the first parameter, thereby improving the ease of judging whether the voice transmitted from the voice transmission system 100 is an artificial voice.

[0077] 3-3. Third Variation (Voice Transmission as Dialogue)

[0078] In the above embodiment, the artificial voice transmitted to the user by the voice transmission system 100 is an artificial voice that transmits voice unidirectionally for information notification, etc. However, the artificial voice transmitted by the voice transmission system 100 may be an artificial voice for the purpose of dialogue with the user.

[0079] Figure 6 3 shows an example of processing performed by the voice transmission program when transmitting voice for the purpose of conversation with the user. In this variant, the voice transmission system 100 includes a microphone 50 for collecting the voice uttered by the user. When the processor 11 executes the process 311, the user's voice is collected by the microphone 50, and the collected voice is analyzed. The processor 11 can also use a machine learning model to perform voice analysis.

[0080] Next, the processor 11 executes process 312. In process 312, the content of the dialogue for responding to the user's voice analyzed in process 311 is determined by the machine learning model.

[0081] Process 313, process 314 and process 315 and Figure 3The processes 113, 114 and 115 are the same. The artificial voice based on the content of the conversation determined by the process 312 is generated by the process 315 and transmitted to the user through the speaker 30.

[0082] When the artificial voice is set as a voice for the purpose of conversation, there is a high possibility that the voice transmission implemented by the voice transmission system 100 continues for a certain period of time. Therefore, the user can easily perceive the change of the second parameter of the artificial voice, thereby improving the ease of determining that the voice transmitted from the voice transmission system 100 is artificial voice.

[0083] 3-4. Fourth Modification (Voice Transmission System as an Autonomous Driving System)

[0084] The speech transmission system 100 may also be a part of the automatic driving system of the vehicle 1. In this case, the automatic driving device 200 of the vehicle 1 is constituted by a part or all of the processor 11 and the storage device 12. It should be noted that the automatic driving device 200 may also be a device for automatic driving of the vehicle 1 using a machine learning model.

[0085] 3-5. Fifth variation (upper and lower limits of the second parameter)

[0086] Alternatively, an upper limit and a lower limit may be set for the second parameter. By setting the upper limit and the lower limit, even if the first parameter changes significantly, the second parameter can be prevented from changing significantly too much. In particular, in the case where the second parameter is the size or height of the artificial voice, setting the upper limit and the lower limit is effective. That is, in the case where the second parameter is the size of the artificial voice, by setting the upper limit and the lower limit for the second parameter, the size of the artificial voice can be prevented from becoming excessively larger or smaller in conjunction with the first parameter, thereby making it difficult for the user to hear clearly. Similarly, in the case where the second parameter is the height of the artificial voice, the height of the artificial voice can be prevented from becoming excessively higher or lower, thereby making it difficult for the user to hear clearly.

[0087] The upper limit and lower limit of the second parameter are set in advance in consideration of the ease of hearing of the user, etc., and are stored in the storage device 12. It should be noted that, when the second parameter exceeds the upper limit or is less than the lower limit, the second parameter will not change even if the first parameter changes further significantly, but in this case, since the nature of the artificial voice changes to a level that is unnatural as a human voice, it is considered that the user is unlikely to mistakenly recognize the voice generated by the voice transmission system 100 as a human voice.

[0088] 4. Application examples

[0089] 4-1. First example

[0090] Finally, a specific example of a scenario in which the voice transmission system 100 of the present embodiment is applied is described. In the first example, a scenario in which the vehicle 1 is traveling on a laterally inclined road is considered. In addition, the first parameter is set to the roll angle of the vehicle 1, and the second parameter is set to the sound quality of the artificial voice. Assuming that the vehicle 1 is likely to roll over due to the excessive speed of the vehicle 1 and the roll angle becomes large, a voice for warning the driver is transmitted through the voice transmission system 100. At this time, according to the change in the sound quality of the artificial voice linked to the roll angle of the vehicle 1, the driver can easily recognize that the voice transmitted from the voice transmission system 100 is an artificial voice. At the same time, by changing the sound quality of the artificial voice in conjunction with the size of the roll angle of the vehicle 1, the driver can also feel the size of the roll angle through the change in sound quality, thereby easily recognizing what degree of operation is required to restore the correct vehicle posture.

[0091] 4-2. Second example

[0092] As a second example, consider that the vehicle 1 is a vehicle towing a trailer, and the identification sensor 22 includes a rear camera that shoots the rear of the vehicle 1. In this case, the first parameter may be set to a parameter associated with the swing of the trailer towed by the vehicle 1, and the second parameter may be set to the size of the artificial voice. For example, the speed, size, period, etc. of the swing are set as the first parameter as parameters associated with the swing of the trailer. It should be noted that the processor 11 can obtain these parameters associated with the swing of the trailer through image analysis of the image captured by the rear camera.

[0093] Assume that when the trailer towed by the vehicle 1 begins to swing left and right, the voice transmission system 100 transmits an artificial voice to the user to inform that the trailer is swinging. At this time, for example, the volume of the artificial voice is linked to the size of the swing, such that the larger the swing of the trailer, the larger the artificial voice. For the user, it becomes easy to distinguish that the voice transmitted from the voice transmission system 100 is an artificial voice through the change of the second parameter linked to the first parameter. At the same time, the size of the swing is easy to identify through the size of the artificial voice. That is, the larger the swing and the more important the response, the louder the volume of the artificial voice, so it is also possible to obtain an effect that it is easier for the user to grasp the importance of the situation notified by the artificial voice.

Claims

1. A voice transmission system, characterized in that: The speech delivery system comprises one or more processors, The one or more processors are configured to: Generate artificial speech that imitates human speech through machine learning models; delivering the artificial voice to a user of the vehicle; as well as A second parameter related to the property of the artificial voice is changed in conjunction with a first parameter related to at least one of the movement of the vehicle and the operation amount of the driver of the vehicle.

2. The speech transmission system according to claim 1, characterized in that The first parameter is any one of a speed, an acceleration, and a steering angle of the vehicle.

3. The speech transmission system according to claim 1 or 2, characterized in that: The second parameter includes at least any one of the size, height, and sound quality of the artificial voice.

4. The speech transmission system according to claim 3, characterized in that: The first parameter is the acceleration of the vehicle, The second parameter includes the height of the artificial voice, The greater the acceleration, the higher the height of the artificial voice caused by the machine learning model.

5. The speech transmission system according to claim 3, characterized in that: The second parameter is provided with an upper limit and a lower limit.

6. The speech transmission system according to claim 1, characterized in that: The first parameter is a difference between an ideal operation amount and an operation amount actually input by the driver.

7. A voice transmission method, characterized in that: The following steps are involved: Generate artificial speech that imitates human speech through machine learning models; delivering the artificial speech to a user of the vehicle; as well as The parameter related to the property of the artificial voice is changed in conjunction with a parameter related to at least one of the movement of the vehicle and the operation amount of the driver of the vehicle.

8. A computer-readable non-transitory storage medium storing a speech delivery program executed by a computer, wherein the computer-readable non-transitory storage medium is characterized in that: The speech delivery program causes the computer to perform the following steps: Generate artificial speech that imitates human speech through machine learning models; delivering the artificial speech to a user of the vehicle; and The parameter related to the property of the artificial voice is changed in conjunction with a parameter related to at least one of the movement of the vehicle and the operation amount of the driver of the vehicle.

Images