Control of an audio content playback function in a vehicle

Abstract

The invention relates to a method for implementing an audio content rendering function in a motor vehicle infotainment system, the infotainment system comprising a device for controlling the audio content rendering function, and, for at least a first seat of the motor vehicle, a first set of at least one loudspeaker dedicated to the first seat and a first microphone dedicated to the first seat. The method comprises receiving (203) a first voice instruction from a first user seated on the first seat, by the first microphone and obtaining (204), by applying a first linguistic generative model to the first voice instruction, a first audio content. Then, the control device can control (205) the first set for the rendering of the first audio content. FIG. 2

Description

Title of the invention: Control of an audio content reproduction function in a vehicle

[0001] The present invention belongs to the field of vehicle infotainment systems, in particular for the implementation of a function for reproducing personalized audio content for one or more passengers of the vehicle.

[0002] The term “vehicle” means any type of vehicle such as a private, utility or heavy goods vehicle, for example.

[0003] Vehicles are now equipped with infotainment systems allowing passengers, in particular those other than the driver, to access multimedia content displayed on one or more screens in the vehicle, and / or reproduced in sound form on a set of speakers.

[0004] However, user interfaces for accessing such multimedia content are standardized and cannot currently be customized to the users seated in the vehicle, nor can they be differentiated between these users.

[0005] In addition, the selection of multimedia content to be rendered requires an active search on the part of the vehicle passengers, a search which may not be easy to implement depending on the user interface of the infotainment system.

[0006] Furthermore, it is desirable to allow a vehicle user, for example a passenger other than the driver, to access multimedia content without disturbing other passengers in the vehicle and / or to allow users to access different content without disrupting the user experience of each of them.

[0007] There is thus a need to improve the user experience of vehicle passengers provided by a vehicle infotainment system, in particular by improving access to targeted multimedia content, by improving its reproduction and / or by allowing vehicle passengers to access distinct multimedia content while allowing a good user experience for each of them.

[0008] The present invention improves the situation.

[0009] To this end, a first aspect of the invention relates to a method for implementing an audio content playback function in a motor vehicle infotainment system, the infotainment system comprising a device for controlling the audio content playback function, and, for at least one first seat of the motor vehicle, a first set of at least one loudspeaker dedicated to the first seat and a first microphone dedicated to the first seat. The method comprises the following steps: - receiving a first voice instruction from a first user sitting on the first place, by the first microphone; - obtaining, by applying a first linguistic generative model to the first voice instruction, a first audio content; - control of the first set of at least one loudspeaker for the reproduction of the first audio content.

[0010] Thus, the generation of audio content takes advantage of the personalization capabilities offered by artificial intelligence, in particular by generative linguistic models of the LLM type, for “Large Language Model” in English. The audio content rendered is thus customizable according to the voice instruction received from the user. In addition, the user interface, namely the microphone, and the set of at least one loudspeaker, are dedicated to the seat on which the user is seated, which makes it possible to restore personalized audio content only to the requesting user, without disturbing the other passengers.

[0011] According to embodiments, the infotainment system may further comprise at least a first avatar dedicated to the first place, the first avatar being a digital avatar displayed on a first screen dedicated to the first place or being a real object dedicated to the first place, the control device being able to control a state of said first avatar, and the method may further comprise the following steps: - controlling the avatar in a first state to indicate to a user seated in the first place to pronounce a voice instruction; - control of the avatar in a second state to indicate the playback of the first audio content.

[0012] Thus, the user interface further comprises an avatar for guiding the user in the use of the audio content playback function, which improves the user experience and facilitates obtaining relevant audio content.

[0013] According to embodiments, the method may further comprise selecting a first user profile by the first user, and the first audio content may further depend on the selected first user profile.

[0014] Thus, the personalization of audio content is improved, because it depends not only on the voice instruction but also on a user profile.

[0015] Additionally, the first avatar may be a digital avatar and a general appearance of the first avatar may depend on the first selected user profile.

[0016] Thus, the user interface associated with the audio content rendering function can also be customized, in addition to the customization of the rendered audio content. Interaction with the user is thus facilitated, which improves the user experience.

[0017] According to embodiments, the infotainment system may further comprise, for at least a second seat of the motor vehicle, a second set of at least one loudspeaker dedicated to the second place and a second microphone dedicated to the second place, and the method may further comprise the following steps: - receiving a second voice instruction from a second user seated in the second seat, through the second microphone; - obtaining, by applying the first linguistic generative model, or another linguistic generative model, to the second voice instruction, a second audio content; - control of the second set of at least one speaker for the reproduction of the second audio content.

[0018] Thus, the audio content rendering function is implemented independently for the first user and for the second user. Each comprises a dedicated user interface, comprising the microphone and a dedicated speaker assembly for rendering the first and second audio contents separately.

[0019] According to embodiments, the method may further comprise, prior to receiving the first voice instruction, controlling the first set of at least one speaker to produce an audio message indicating to the first user that they can request audio content.

[0020] Thus, the use of the audio content rendering function is made easier for the first user.

[0021] A second aspect of the invention relates to a computer program comprising instructions for implementing the method according to the first aspect of the invention, when these instructions are executed by a processor.

[0022] A third aspect of the invention relates to an infotainment system for a motor vehicle comprising a control device configured to implement an audio content playback function for at least a first seat of the motor vehicle, the infotainment system comprising: - a first set of at least one loudspeaker dedicated to the first place; - a first microphone dedicated to the first place; wherein said control device is configured to receive a voice instruction from a user seated in the first place and picked up by the first microphone, to obtain, by applying a linguistic generative model to the voice instruction, a first audio content, and to control the first assembly for the reproduction of the first audio content.

[0023] According to embodiments, the first microphone may be a directional microphone arranged in a seat facing the first position.

[0024] Such an embodiment makes it possible to capture the first voice instruction without requiring a high sound level from the first user. The interaction of the The first user with the infotainment system is thus done without disturbing the other passengers in the vehicle.

[0025] In addition or as a variant, the first assembly may comprise a first loudspeaker and a second loudspeaker arranged on either side of a headrest of a seat in the first place.

[0026] Thus, an immersive sound bubble is formed around the first user which improves the user experience associated with the restitution of the audio content.

[0027] Other characteristics and advantages of the invention will appear on examining the detailed description below, and the appended drawings in which:

[0028] [Fig.l] illustrates a motor vehicle according to embodiments of the invention;

[0029] [Fig.2] is a diagram illustrating the steps of a method for controlling a navigation function according to embodiments of the invention;

[0030] [Fig.3] illustrates a control device of an infotainment system of a motor vehicle, according to embodiments of the invention.

[0031] [Fig.l] illustrates a motor vehicle 100, according to embodiments of the invention

[0032] The vehicle 100 comprises in particular a control device 101, which may be a centralized control device in charge of a plurality of functions of the motor vehicle. The control device 101 may be of the ECU type in particular, for “Electronic Control Unit” in English.

[0033] According to the invention, the control device 101 can be responsible for controlling at least one function of an infotainment system of the vehicle 100.

[0034] The control device 101 is notably configured to implement a function for rendering audio or audio-visual content according to the invention. In the following, the term “audio content” includes both content rendered solely in sound form, but also the audio components of multimedia content further comprising a visual component displayed on a screen.

[0035] The audio content may be derived from a linguistic generative model derived from artificial intelligence, such as a large language model, also called LLM for "Large Language Model" in English, trained on a corpus of texts used as training data to build the linguistic generative model.

[0036] Such a generative model is configured to receive as input an instruction or "prompt", resulting from a user's voice instruction, comprising a natural language sentence or keywords, and to produce natural language audio content as output, the audio content being determined based on the instruction. Alternatively, the generative linguistic model is configured to generate text content which is then converted into audio content to be played back audibly by the vehicle's infotainment system 100.

[0037] No restriction is attached to the structure of the linguistic generative model, which may be an algorithm implementing an artificial neural network, such as a deep neural network.

[0038] The linguistic generative model may be stored locally in the vehicle 100, in an internal memory of the control device 101, or in a memory 102 of the vehicle 100. The control device 101 may thus locally execute the linguistic generative model to apply it to a voice instruction received from a passenger of the vehicle, or a part of such a voice instruction, to obtain audio content as output.

[0039] Alternatively, the linguistic generative model is stored in a remote server, which the control device 101 accesses via a communication interface 103, which may be a cellular interface for accessing a mobile cellular network, such as a 3G, 4G, 5G network or any other mobile cellular network. In this alternative, the control device 101 may transmit a request to the remote server, the request comprising the voice instruction or a part of the voice instruction, and receives in return audio content generated by the linguistic generative model executed in the remote server.

[0040] According to the invention, the infotainment system of the vehicle 100 comprises, for at least one given seat of the vehicle 100: - a set of at least one loudspeaker dedicated to the given place of the vehicle, for example a set of two loudspeakers located near the given place, for example at the height, and on either side, of a headrest of the seat of the given place, thus forming an immersive sound bubble around a user seated in the given place. For example, the set of loudspeakers comprises two loudspeakers located less than 50 centimeters from the given place, in particular less than 50 cm from the ears of a user positioned in the given place, and preferably less than 30 centimeters. The loudspeakers of the set may be capable of being controlled by the control device 101 to reproduce two components of the same audio content, in stereo for example; - a microphone dedicated to the given place, which may be a directional microphone, so as to pick up a voice instruction, including a low voice, from the user located in the given place; and - an avatar representing the audio content playback function, or more generally the infotainment system, to the user positioned in the given seat. The avatar may be a physical avatar, or may be a virtual avatar displayed on a screen located near the given seat, for example in the back of the seat located in front of the given seat, when the given seat is in the rear of the vehicle 100. The avatar thus forms an element of the user interface of the audio content playback function. audio content production.

[0041] Thus, the user can communicate with the infotainment system and receive audio content without disturbing other users of the vehicle 100.

[0042] Preferably according to embodiments, the infotainment system may comprise the above list of elements for several seats of the vehicle, for example for a left rear seat 110.1 of the vehicle 100 and for a right rear seat 110.2 of the vehicle 100.

[0043] According to these embodiments, the infotainment system therefore comprises, near the left rear seat 110.1: - a set 111.1 of at least one loudspeaker, for example two loudspeakers located on either side of the left rear seat 110.1, in particular to the right and left of the head of the user seated on the left rear seat; - a 112.1 microphone, which can be positioned in front of the left rear seat 110.1, for example in the rear of the left front passenger seat, which may be the driver. Microphone 112.1 is preferably a directional microphone; - a real or virtual avatar 113.1 displayed on a screen that can be positioned opposite the left rear seat 110.1, for example in the back of the left front passenger seat. Thus, when the avatar is virtual, the reference 113.1 is assigned to the screen on which the virtual avatar is displayed.

[0044] The infotainment system according to these embodiments further comprises, near the right rear seat 110.2: - a set 111.2 of at least one loudspeaker, for example two loudspeakers located on either side of the right rear seat 110.2, in particular to the right and left of the head of the user positioned on the right rear seat 110.2; - a 112.2 microphone, which can be positioned in front of the left rear seat 110.2, for example in the back of the right front passenger seat. The microphone 112.2 is preferably a directional microphone; - a real or virtual avatar 113.2 displayed on a screen that can be positioned opposite the right rear seat 110.2, for example in the back of the right front passenger seat. Thus, when the avatar is virtual, the reference 113.2 is assigned to the screen on which the virtual avatar is displayed.

[0045] Thus, users seated in the rear of the vehicle, who may be children, can interact independently of one another, without disturbing each other, with the infotainment system of the vehicle 100 to obtain a reproduction of audio content that is different from one another.

[0046] An avatar according to the invention is capable of representing the audio content rendering function according to the invention. The avatar can thus guide the user located near the avatar for the formulation of a voice instruction from the user, voice instruction which is used by the control device 101 to obtain audio content from the linguistic generative model, and aims to give the impression that it vocally renders the audio content once obtained.

[0047] For this purpose, the avatar can have several states, visually distinguishable by the user sitting near the avatar. The states of the avatar are controlled by the control device 101 of the infotainment system.

[0048] The avatar may be in a first state indicating that the avatar is listening, when a voice instruction is expected from the user in the vicinity of the avatar. The first listening state may further indicate information on the sound level picked up by the microphone, so that the user can adjust the sound level of his voice. Preferably, the information indicates by color whether the sound level is adequate, too loud or too soft. An adequate level that is intermediate between a level that is too soft and too loud ensures that the user's voice instruction can be correctly picked up by the microphone, and interpreted by the control device 101, without however disturbing the other users of the vehicle 100.

[0049] The avatar may be in a second state indicating speech by the avatar, during which sound data is played back on the set of speakers, the sound data being able to be messages guiding the user or being able to be the audio content during its playback.

[0050] The avatar may further be in a third state indicating a misunderstanding of a voice instruction received from the user, and thus inviting the user to reformulate his voice instruction.

[0051] The avatar, in particular the sequential adoption of the states described previously under the control of the control device 101, makes it possible to assist the user in his interactions with the audio content reproduction function.

[0052] Avatar states may differ: - by a displayed color; and / or - by the appearance, in particular the shape, of the avatar.

[0053] In an example of an avatar that is a real physical object, the avatar has a general appearance of an animal, for example a rabbit. A general appearance is a two-dimensional or three-dimensional constitution that can be broken down into several states. In the example of the rabbit, the transition from one state to another can be characterized by a variation in the position of the ears and mouth of the avatar, to alternately indicate a listening posture or a speaking posture.

[0054] However, there are no restrictions on the general appearance of the avatar, the number of states of the avatar, or the manner in which each state is represented on the avatar.

[0055] Each avatar may depend on a profile of the user located near the avatar. For example, a first user sitting in the left rear seat 110.1 may be associated with a child profile, as identified to the infotainment system as being between 5 and 12 years old, while a second user sitting in the right rear seat 110.2 may be associated with a teenager profile, as identified to the infotainment system as being between 13 and 17 years old. Other possible profiles could be an adult profile or a young child profile for example.

[0056] A first avatar associated with the child profile of the first user can thus have a first general appearance different from a second general appearance of a second avatar associated with the adolescent profile of the second user.

[0057] When the first and second avatars are virtual, the avatar can change its general appearance for the same given place, when the user sitting on the given place changes profile, or when another user associated with another user profile sits on the given place.

[0058] [Fig.2] is a diagram illustrating the steps of a method for implementing an audio content rendering function by a vehicle infotainment system, according to embodiments of the invention.

[0059] In a step 200, the audio content rendering function is initiated by the control device 101, for at least one given seat of the vehicle. In the following, for illustrative purposes only, it is considered that the audio content rendering function is initiated by the control device 101 for the left rear seat 110.1 on which a user is seated. The initiation may follow the selection of a user profile by the user, for example the child profile mentioned above. The selection of the user profile may be made upon receipt of a user input on a user interface. The user input may be a voice input picked up by the microphone 112.1. The voice input may be linked to the audio content rendering function by detecting one or more keywords in the voice input, for example by detecting a name of the service associated with the audio content rendering function.

[0060] The initiation can thus comprise the execution of the audio content rendering function, which, once executed, is in an initial state ready to interact with the user.

[0061] The audio content rendering function can be adapted to the user profile. In particular, the audio content generated and then rendered can be adapted to the user profile: for the child profile, considered in the following for illustrative purposes, the audio content rendering function can be dedicated to the audio rendering of children's stories. Alternatively, for an adult profile for example, the audio content rendering function can be dedicated to informative audio content, according to the user's instructions, such as the audio rendering of a historical summary on an ongoing conflict in the world, the audio rendering of a definition of a scientific theory or even the sound reproduction of comments or information on a city close to the vehicle 100 or close to a destination of the vehicle 100, indicated by the driver in a navigation function not shown or described in this description.

[0062] For this purpose, several linguistic generative models can be provided, each linguistic generative model being dedicated to each profile. Alternatively, a single generic linguistic generative model is used, and the voice instruction is enriched with information identifying the user profile, as input to the linguistic generative model.

[0063] Thus, no restriction is attached to the type of audio content rendered according to the invention, which can advantageously depend on a user profile identified by the user seated in the given seat for which the function is initiated.

[0064] Following the initiation of step 200, the control device 101 commands the activation of the avatar associated with the audio content rendering function, at a step 201.

[0065] In the case of a virtual avatar, activation of the avatar includes displaying the avatar on the screen 113.1.

[0066] In the case of an actual avatar 113.1, the activation of the avatar may comprise the activation of a light indicator or a welcome animation, indicating to the user that the avatar is active.

[0067] Following step 202, the control device 101 controls the avatar so that the avatar is in the first speech state of the avatar, and controls the set 111.1 of speakers to produce a first audio message to the user indicating that he has the possibility of requesting the reproduction of audio content. For example, the control of the set 111.1 may consist of producing in an audible manner a first message such as "Hello, do you want me to tell you a story?".

[0068] Alternatively, the first message may include a suggestion or an example of a voice instruction, in order to guide the user and facilitate the choice. For example, the first message may be “Hello, do you want me to tell you a story? You can ask me for a story of your choice, for example a story with a unicorn and a little girl.”

[0069] The first message may depend on the user profile selected during step 200. The first message may thus be adapted depending on whether the user is a child or an adult.

[0070] Following the command of the avatar in the first state, and the sound production of the first message during step 202, the control device 101 commands the avatar 113.1 at a step 203 so that the avatar 113.1 is in the second listening state, thus indicating to the user that the infotainment system is waiting for a voice instruction from the user seated in the seat 110.1.

[0071] In step 203, the control device 101 may also activate a timer, ini- set to a given duration, for example a few seconds.

[0072] If a voice instruction is received from the user during step 203, the method proceeds to step 204 described below. If a voice instruction is not received during step 203 upon expiration of the timer, or if a sentence spoken by the user cannot be interpreted as a voice instruction, the method proceeds to step 206 described below. For example, a sentence such as "I don't know" spoken in step 203 is considered by the control device 101 not to be a voice instruction, and the method proceeds to step 206.

[0073] In step 204, an audio content is obtained by the control device 101, from the received voice instruction, and by execution, locally or remotely, of the aforementioned linguistic generative model providing the audio content as output.

[0074] For example, in the situation in which the user is associated with a child profile, and in which the linguistic generative model generates a story for a child, the voice instruction may be "can you tell me the story of a unicorn who has a sore paw and who meets a little girl?". In this situation, the linguistic generative model is trained beforehand to generate a story for a child from a voice instruction indicating contextual elements from which to generate the story. The aforementioned example of voice instruction may thus be submitted in its entirety as input to the linguistic generative model, which is capable of extracting keywords, or prior processing makes it possible to extract the keywords therefrom, to submit the keywords thus extracted as input to the linguistic generative model.

[0075] At a step 205, the control device 101 controls the set 111.1 of loudspeakers so as to restore the sound content to the child seated in the rear left seat 110.1, and controls the avatar in the first speech state.

[0076] When, in step 203, no voice instruction is received, or when the spoken sentence cannot be interpreted by the control device 101 as a voice instruction, the control device 101 may control the avatar so that the avatar is in the third state of incomprehension, and may control the set 111.1 of loudspeakers to produce a second sound message to the user. The second sound message may be identical to the first sound message, or may give additional information compared to the first sound message, in order to assist the user in the selection and formulation of the voice instruction. The additional information may comprise an example of a voice instruction.For example, if the first message only states "Do you want me to tell you a story?", the second message may state "You can, for example, ask me to tell you a story about a magical animal that has an encounter." The method then returns to step 203, again putting the avatar into the second listening state, to receive a voice instruction from the user.

[0077] [Fig.3] shows the structure of a control device 101 according to embodiments of the invention.

[0078] The control device 101 comprises a processor 301 configured to communicate unidirectionally or bidirectionally, via one or more buses or via a direct wired connection, with a memory 302 such as a memory of the “Random Access Memory” type, RAM, or a memory of the “Read Ordy Memory” type, ROM, or any other type of memory (Flash, EEPROM, etc.). Alternatively, the memory 302 comprises several memories of the aforementioned types.

[0079] The memory 302 is capable of storing, permanently or temporarily, at least some of the data used and / or resulting from the implementation of the method described with reference to [Fig.2].

[0080] In particular, the memory 302 may store the first and second messages, and may store the linguistic generative model, or several linguistic generative models.

[0081] The processor 301 is capable of executing instructions, stored in the memory 302, for implementing the steps of the method according to the invention, described with reference to [Fig. 2]. Alternatively, the processor 301 can be replaced by a microcontroller designed and configured to carry out the steps of the method according to the invention, described with reference to [Fig. 2].

[0082] The control device 101 comprises a first control interface 303 capable of controlling each avatar 113.1-113.2 of the vehicle 100. If the avatar is a real avatar, the first control interface 303 is capable of controlling the state of the real avatar. If the avatar is displayed on a screen, the first control interface 303 is capable of controlling the display of the avatar on the screen in a given state. When the vehicle comprises several avatars or several screens, the control device 101 may comprise a first control interface 303 for each avatar or for each screen.

[0083] The control device 101 comprises a second control interface 304 capable of controlling each set 111.1-111.2 of speakers of the vehicle 100, to reproduce the audio content, or broadcast the first message or the second message in particular. When the vehicle comprises several avatars or several screens, the control device 101 may comprise a second control interface 304 for each avatar or for each screen.

[0084] The control device 101 further comprises a reception interface 305 capable of receiving data captured by each microphone 112.1-112.2 of the vehicle 100. When the vehicle 100 comprises several microphones, the control device 101 may comprise a reception interface 305 for each microphone.

[0085] The control device 101 may further comprise an access interface 306 to the vehicle memory 102 described previously, which can store the linguistic generative model.

[0086] The control device 101 may further comprise an interface 307 capable of communicating bidirectionally with the communication interface 103 of the vehicle, to access a mobile telecommunications network.

[0087] The control device 101 may comprise other interfaces for communicating with other equipment of the vehicle 100.

[0088] The present invention is not limited to the embodiments described above as examples; it extends to other variants.

Claims

Claims

1. Method for implementing an audio content rendering function in a motor vehicle infotainment system (100), the infotainment system comprising a control device (101) for the audio content rendering function, and, for at least a first seat (110.1) of the motor vehicle, a first set (111.1) of at least one loudspeaker dedicated to the first seat and a first microphone (112.1) dedicated to the first seat, in which the method comprises the following steps: - receiving (203) a first voice instruction from a first user seated on the first seat, by the first microphone; - obtaining (204), by applying a first linguistic generative model to the first voice instruction, a first audio content; - controlling (205) the first set of at least one loudspeaker for the rendering of the first audio content.

2. The method of claim 1, wherein the infotainment system further comprises at least one first avatar (113.1) dedicated to the first place, the first avatar being a digital avatar displayed on a first screen dedicated to the first place or being a real object dedicated to the first place, the control device (101) being capable of controlling a state of said first avatar, and the method further comprising the following steps: - controlling (202) the avatar in a first state to indicate to a user seated in the first place to pronounce a voice instruction; - controlling (205) the avatar in a second state to indicate the reproduction of the first audio content.

3. The method of claim 1 or 2, further comprising selecting (200) a first user profile by the first user, and wherein the first audio content is further dependent on the selected first user profile.

4. A method according to claim 2 and claim 3, wherein the first avatar (113.1) is a digital avatar and wherein a general appearance of the first avatar depends on the selected first user profile.

5. Method according to one of the preceding claims, in which the infotainment system further comprises, for at least one second seat (110.2) of the motor vehicle (100), a second set (111.2) of at least one loudspeaker dedicated to the second place and a second microphone (112.2) dedicated to the second place, and the method further comprising the following steps: - receiving (203) a second voice instruction from a second user seated on the second place, by the second microphone; - obtaining (204), by applying the first linguistic generative model, or another linguistic generative model, to the second voice instruction, a second audio content; - controlling (205) the second set of at least one loudspeaker for the reproduction of the second audio content.

6. Method according to one of the preceding claims, further comprising, prior to receiving (203) the first voice instruction, controlling (202) the first set (111.1) of at least one loudspeaker to produce an audio message indicating to the first user that he can request audio content.

7. Computer program comprising instructions for implementing the method according to one of the preceding claims, when these instructions are executed by a processor (301).

8. Infotainment system for a motor vehicle comprising a control device (101) configured to implement an audio content rendering function for at least a first seat (110.1) of the motor vehicle, the infotainment system comprising: - a first set (111.1) of at least one speaker dedicated to the first seat; - a first microphone (112.1) dedicated to the first seat; wherein said control device is configured to receive a voice instruction from a user seated on the first seat and picked up by the first microphone, to obtain, by applying a linguistic generative model to the voice instruction, a first audio content, and to control the first set for the rendering of the first audio content.

9. An infotainment system according to claim 8, wherein the first microphone (112.1) is a directional microphone arranged in a seat facing the first position (110.1).

10. An infotainment system according to claim 8 or 9, wherein the first assembly (111.1) comprises a first speaker and a second speaker arranged on either side of a headrest of a first place seat (110.1).

Images