Method for controlling a sound diffusion system in the passenger compartment of a vehicle for messages relating to explanatory information about ADAS functions or automated driving of the vehicle.

Abstract

The present invention relates to a method and device for controlling a sound system in the passenger compartment of a vehicle, providing information explaining the vehicle's driver assistance functions. The method (51, 52, 53) obtains a third piece of data from a first text-type data point representing a question asked by a vehicle occupant and relating to information explaining one of said driver assistance or automated driving functions, and a second text-type data point relating to the current driving context of the vehicle. The method (54, 55) obtains a fourth piece of data point representing a response relating to the information explaining the driver assistance or automated driving functions corresponding to the third piece of data point and controls (56) the diffusion of the fourth piece of data in the vehicle's passenger compartment. Figure 5

Description

Title of the invention: Method for controlling a sound diffusion system in a vehicle passenger compartment for messages relating to explanatory information about AD AS functions or automated driving of the vehicle. technical field

[0001] The present invention relates to the control of a sound diffusion system, in a vehicle's passenger compartment, of a message relating to explanatory information of vehicle driving assistance functions or automated vehicle driving. Technological background

[0002] Modern vehicles incorporate various driver assistance systems, known as ADAS (Advanced Driver-Assistance System) or automated driving systems, to assist the driver of such a vehicle in controlling the driving of the vehicle, or even to replace the driver in an autonomous driving mode of the vehicle.

[0003] In most cases, ADAS functions, i.e., driver assistance functions implemented by ADAS systems, and automated driving functions, i.e., those implemented by automated driving systems, are not easy for some drivers to understand and manage, particularly in demanding driving situations. This results in a lack of confidence, or even distrust, among these drivers in ADAS and automated driving systems, primarily due to a poor understanding of these systems and their influence on vehicle behavior. Therefore, raising awareness among these drivers about the operation and benefits of ADAS and automated driving functions is essential for ADAS and automated driving systems to be adopted and used correctly.

[0004] In order for a driver of a vehicle to become aware of the operation and advantages of AD AS functions and automated driving functions implemented in that vehicle, it may be useful to provide explanatory information on these AD AS and automated driving functions in real traffic conditions, i.e. when the driver is driving that vehicle.

[0005] Solutions exist for providing this explanatory information under real-world driving conditions. These solutions interact with the driver but are not contextualized, meaning they do not take into account the context in which The vehicle's journey is unfolding, and these solutions are neither personalized, meaning they do not take into account the driver or other vehicle occupants. Existing solutions provide the same explanatory information as soon as a vehicle occupant requests it, regardless of the context. Summary of the present invention

[0006] One object of the present invention is to solve at least one of the problems of the technological background described above.

[0007] According to a first aspect, the present invention relates to a method for controlling a sound diffusion system in a vehicle's passenger compartment, for messages relating to explanatory information about vehicle driver assistance or automated vehicle driving functions, said method being implemented by at least one processor and comprising the following steps: - obtaining, via a voice recognition system, a first text-type data representative of a question asked by a vehicle occupant and relating to explanatory information of one of the said vehicle driving assistance or automated vehicle driving functions; - obtaining, from at least one contextual data capture system, a second text-type data item relating to a current vehicle driving context; - obtaining, from a text-type data fusion system, a third text-type data item from the first and second data items, both of which are text-type; - submission of the third data to a conversational system based on a language model having access to a database including data representative of the explanatory information of the vehicle's driving assistance or automated driving functions; - obtaining, from the conversational system, a fourth piece of data representing a response relating to the explanatory information of the vehicle driving assistance or automated vehicle driving functions from the database corresponding to the third piece of data; - control of the sound diffusion system for broadcasting the fourth data point in the vehicle's passenger compartment.

[0008] The implementation of the method makes it possible to create a virtual companion for the driver which teaches the driver how to optimally use the AD AS and vehicle automation functions according to the current driving context of the vehicle.

[0009] The method offers the driver an educational course on how AD AS and automated driving functions can benefit the comfort of the driver and vehicle occupants and on how AD AS and Automated driving systems can provide advance warnings / automated maneuvers for driving situations that may potentially be risky.

[0010] But the process does not tell the driver what to do because the responsibility for the actions and their effects lies with the driver.

[0011] The conversational system is based on a language model that has access to a database containing explanatory information about the AD AS and automated driving functions. Such a system thus implements a question / answer system with fluid interactions with the driver.

[0012] The method is empathetic, meaning that it provides explanatory information about ADAS and automated driving functions that is personalized according to the vehicle's current context. The vehicle's driver can thus be educated without obligation or constraint, which contributes to the driver's adoption of ADAS and automated driving systems. For example, interaction with the method can be initiated as soon as an ADAS or automated driving function is activated. Explanatory information relating to this ADAS function, such as scenario / context information, can then be displayed in the vehicle's cabin when the function is activated to improve the explanation of the ADAS or automated driving function's behavior, the reason for its activation, and how the driver can best use it.

[0013] Furthermore, the use of data representative of a current vehicle context makes it possible to provide the most appropriate educational interactions with the driver at any time and in any place when the latter is not occupied with other tasks.

[0014] According to a particular and non-limiting embodiment of the present invention, the second data comprises at least one word or sequence of words from the natural language representative of a context external to the vehicle, said external context being obtained from a current geographical position of the vehicle.

[0015] According to a particular and non-limiting embodiment of the present invention, the external context is defined by one of the following data or one of their combinations; - a data point representative of a driving scenario; - representative data of road traffic and environmental conditions.

[0016] The method is then advantageous because the explanatory information of vehicle driving assistance or automated vehicle driving functions is chosen and disseminated according to the current external context of the vehicle, i.e. according to the current driving scenario, the current road environment around the vehicle and the current road traffic and environmental conditions.

[0017] According to a particular and non-limiting embodiment of the present invention, the second data element comprises a word or a sequence of words from the natural language representative of an internal context of the vehicle, said internal context being obtained from at least one on-board sensor of the vehicle.

[0018] According to a particular and non-limiting embodiment of the present invention, the internal context is defined by one of the following data or one of their combinations:

[0019] - a data point representing a vehicle speed;

[0020] - a data point representing an acceleration of the vehicle;

[0021] - a representative data point of an activation of one of the assistance functions driving the vehicle or automated driving of the vehicle;

[0022] - a data point representing an emotion of the vehicle driver;

[0023] - a representative data point of the number of occupants of the vehicle;

[0024] - a representative data point of the number of occupants of the vehicle in the process of discuss ;

[0025] - a representative data of a road environment around the vehicle.

[0026] The process is then advantageous because the explanatory information for vehicle driving assistance or automated vehicle driving functions is chosen and disseminated according to the current internal context of the vehicle, i.e. according to the driver's driving attitude and according to the other occupants of the vehicle and their relationship with the driver.

[0027] According to a particular and non-limiting embodiment of the present invention, the lang model comprises a trained deep neural network.

[0028] According to a second aspect, the present invention relates to a control device for a sound diffusion system in a vehicle's passenger compartment, for a message relating to explanatory information for vehicle driving assistance functions or automated vehicle driving, the device comprising a memory associated with a processor configured for implementing the steps of the method according to the first aspect of the present invention.

[0029] According to a third aspect, the present invention relates to a vehicle, for example of the automobile type, comprising a device as described above according to the second aspect of the present invention.

[0030] According to a fourth aspect, the present invention relates to a computer program which includes instructions adapted for carrying out the steps of the process according to the first aspect of the present invention, in particular when the computer program is executed by at least one processor.

[0031] Such a computer program may use any programming language, and be in the form of source code, object code, or an intermediate code between source code and object code, such as in a partially compiled form, or in any other desirable form.

[0032] According to a fifth aspect, the present invention relates to a computer-readable recording medium on which is recorded a computer program comprising instructions for carrying out the steps of the process according to the first aspect of the present invention.

[0033] On the one hand, the recording medium can be any entity or device capable of storing the program. For example, the medium can include a storage means, such as a ROM, RAM, CD-ROM or a microelectronic circuit-type ROM, or a magnetic recording means or a hard disk drive.

[0034] On the other hand, this recording medium can also be a transmissible medium such as an electrical or optical signal, such a signal being able to be transmitted via an electrical or optical cable, by conventional or radio frequency, by self-directing laser beam, or by other means. The computer program according to the present invention can, in particular, be downloaded from an Internet-type network.

[0035] Alternatively, the recording medium may be an integrated circuit in which the computer program is incorporated, the integrated circuit being adapted to execute or to be used in the execution of the process in question. Brief description of the figures

[0036] Other features and advantages of the present invention will become apparent from the description of the specific and non-limiting embodiments of the present invention below, with reference to the attached Figures 1 to 5, in which:

[0037] [Fig. 1] schematically illustrates an implementation environment of the present invention in a vehicle.

[0038] [Fig.2] schematically illustrates a phase of learning the language model, according to a particular and non-limiting example of the present invention.

[0039] [Fig.3] schematically illustrates a deep neural network based on a transformer-type architecture according to a particular and non-limiting embodiment of the present invention.

[0040] [Fig.4] schematically illustrates a device configured to control a sound diffusion system in a vehicle cabin for messages relating to explanatory information on vehicle driving assistance functions or automated vehicle driving, according to a particular and non-limiting embodiment of the present invention.

[0041] [Fig. 5] illustrates a flowchart of the different stages of a method for controlling a sound diffusion system in the passenger compartment of a vehicle, providing information explaining vehicle driver assistance or driving functions. automated vehicle, according to a particular and non-limiting embodiment of the present invention. Description of examples of achievements

[0042] A method and a control device for a sound diffusion system in a vehicle passenger compartment for messages relating to explanatory information about vehicle driver assistance or automated vehicle driving functions, will now be described in what follows with joint reference to Figures 1 to 5. The same elements are identified with the same reference signs throughout the following description.

[0043] The terms "first," "second" (or "firsts," "seconds"), etc., are used in this document by arbitrary convention to allow for the identification and distinction of different elements (such as operations, means, etc.) implemented in the embodiments described below. Such elements may be distinct or correspond to a single element, depending on the embodiment.

[0044] A vehicle, according to the present invention, corresponds, for example, to a vehicle with an internal combustion engine, with electric motor(s), or even a hybrid vehicle with an internal combustion engine and one or more electric motors. The vehicle thus corresponds, for example, to a land vehicle, for example a car, a truck, a bus.

[0045] Fig. 1 schematically illustrates an implementation environment of the present invention in a vehicle.

[0046] The vehicle includes a device 2 of the [Fig.4].

[0047] The vehicle includes a voice recognition system 10, a system of contextual data capture 11, a conversational system 12, a text-type data fusion system 13 and a sound diffusion system 14 in the vehicle's interior.

[0048] The speech recognition system 10 is configured to provide a first data 100 of text type representative of a question asked by a driver of the vehicle and relating to explanatory information of one of said AD AS or automated driving functions of the vehicle.

[0049] For example, the speech recognition system 10 includes at least one microphone positioned in the passenger compartment and a computing unit configured to analyze the human voice captured by said at least one microphone in order to transcribe it into the form of the first data 100 of text type which is usable for the text type data fusion system 13.

[0050] This explanatory information regarding the ADAS or automated driving functions of the vehicle is usually that mentioned in a vehicle manual or vehicle maintenance booklet. This information may indicate how an ADAS or automated driving function can be activated or deactivated, explain The effects of these functions, or details concerning the vehicle's behavior when these functions are activated, are stored digitally in the vehicle's database. The database provides access to this information upon request. This request takes the form of a third data entry, 103, of type text, as will be seen later.

[0051] The contextual data capture system 11 is configured to provide a second text data item 110 relating to a current vehicle driving context. This second data item 110 can be used by the text data fusion system 13.

[0052] According to a particular and non-limiting embodiment of the present invention, the second data 110 may include at least one word or sequence of words from the natural language representative of a context external to the vehicle, said external context being obtained from a current geographical position of the vehicle.

[0053] For example, the current geographical position of the vehicle is obtained from a geolocation system or GNSS (Geolocation and Navigation by a Satellite System), for example a GPS type system (from the English "Global Positioning System" or in French "Système de géo-positionnement par satellites") or Galileo, on board the vehicle.

[0054] For example, the external context of the vehicle can be defined by one of the following data or one of their combinations;

[0055] - a data point representative of a driving scenario;

[0056] - data representative of road traffic conditions and environmental.

[0057] The present invention is not limited to these examples of defining an external context of the vehicle nor to the means implemented to obtain a second data defining it, but extends to all means enabling the definition of an external context to a vehicle.

[0058] For example, the words urban, extra-urban, and highways can indicate different driving scenarios. The second data point 100 can then include one of these words.

[0059] For example, the vehicle may include a communication system configured to exchange data with a remote system configured to broadcast weather reports and / or alerts and / or road traffic information. The vehicle's on-board communication system may then query this remote system to obtain data representative of road traffic and environmental conditions. A processing unit in the vehicle may then be configured to translate this data representative of road traffic and environmental conditions into a word or sequence of words in natural language and to add these words or a sequence of words from natural language to the second data 110.

[0060] According to a particular and non-limiting embodiment of the present invention, the second data 110 may include a word or a sequence of words from the natural language representative of an interior context of the vehicle, said interior context being obtained from at least one on-board sensor of the vehicle.

[0061] For example, the internal context is defined by one of the following data or one of their combinations:

[0062] - a data point representing a vehicle speed;

[0063] - a data point representing an acceleration of the vehicle;

[0064] - data representative of an activation of one of the AD AS functions or of automated vehicle driving;

[0065] - a data point representing the emotional state of the vehicle driver;

[0066] - a representative data point of the number of occupants in the vehicle;

[0067] - a representative data point of the number of occupants of the vehicle in the process of discuss ;

[0068] - a representative data of a road environment around the vehicle.

[0069] The present invention is not limited to these examples of defining the vehicle's interior context nor to the means implemented to obtain a second data defining it, but extends to all means enabling the definition of an interior context for a vehicle including all types of sensors and data obtained from these sensors or their combinations.

[0070] For example, a vehicle speed sensor can be adapted to provide a word or sequence of words from natural language that is representative of the vehicle's current speed. The second data point 110 then includes this word or sequence of words from natural language. Basically, a conventional speed sensor can be associated with a processing unit configured to translate the speeds given by the speed sensor into a word or sequence of words. The processing unit can also be configured to translate speeds measured consecutively by the speed sensor into a word or sequence of words representative of a vehicle's acceleration. The second data point 110 then includes this word or sequence of words.

[0071] For example, the vehicle may include a driver emotion detection system configured to detect the driver's emotional state based on an analysis of the driver's face image captured by an on-board camera, in particular the eyes, nose, and lips. The emotion detection system may be configured to provide a natural language word representative of the driver's emotional state, such as anger, disgust, pride, happiness, neutrality, sadness, surprise, etc. The second data point 110 then includes this word.

[0072] For example, the vehicle may include a processing unit configured to count the number of fastened seat belts, i.e., the number of vehicle occupants. This processing unit may then be configured to provide a word or sequence of words from natural language representing this number of occupants. The second data 110 then includes this word or sequence of words from natural language.

[0073] For example, the voice recognition system can be configured to detect whether several occupants of the vehicle are talking and to provide a word or sequence of words indicating whether several occupants are talking and / or how many occupants are talking. The second data point 110 then includes this word or sequence of words.

[0074] For example, the vehicle may include a system for perceiving the external road environment near the vehicle. This system may include LiDAR-type sensors, for example, and / or cameras associated with a processing unit configured to obtain a word or sequence of words representative of the road environment around the vehicle. This data may, for example, be a word describing this environment such as 'fluid', 'congested', 'risky', etc. The second data 110 then includes this word.

[0075] The text-type data fusion system 12 is configured to provide, from the first data 100 and the second data 110, both of text type, a third data 120 representative of a text-type query consisting of a word or a sequence of words.

[0076] For example, the merger consists of concatenating the first data 100 with the second data 110.

[0077] The conversational system 12 is based on a trained language model 131 having access to a database 132 comprising data representative of the explanatory information of the AD AS or automated driving functions of the vehicle.

[0078] The conversational system 12 is configured to provide a fourth data 130 representative of a response relating to the explanatory information of the AD AS or automated driving functions of the vehicle from the database 132 corresponding to the third data 120.

[0079] The sound diffusion system 14 is configured to diffuse the fourth data 130 into the vehicle's passenger compartment.

[0080] For example, the fourth data 130 is audio data. The sound distribution system 14 then includes a processing unit configured to control at least one on-board speaker of the vehicle so that the fourth data is broadcast in the vehicle's passenger compartment.

[0081] For example, if the fourth data 130 is a word or sequence of words from natural language, the broadcasting system may include a speech synthesizer configured to convert this word or sequence of words from natural language into audio data.

[0082] The language model 131 is trained during a learning phase of [Fig.2].

[0083] In a first operation of the learning phase, textual data 202 can be generated as output of the language model 131 to be trained when a set of training textual data 201 is presented as input to the language model 131 to be trained.

[0084] Each textual training data 201 can be a word or a sequence of words from natural language representative, on the one hand, of a question that can be asked by an occupant of the vehicle and relating to explanatory information of one of the said AD AS or automated driving functions of the vehicle and, on the other hand, of a context (interior and / or exterior) of the vehicle.

[0085] Each training text data 201 is associated with an expected text data 212.

[0086] Each text data 202 (generated) or expected (212) can be a word or sequence of words from natural language representative of a response relating to the explanatory information of the AD AS or automated driving functions of the vehicle corresponding to a training text data 201.

[0087] In a second operation of the learning phase, internal parameters of the language model 131 to be trained can be adjusted to minimize a loss function 211 quantifying (210) the differences between the textual data 202 generated as output of the language model 131 to be trained and the expected textual data 212.

[0088] According to a particular and non-limiting embodiment of the present invention, the loss function 211 is of the maximum likelihood type.

[0089] According to a particular and non-limiting embodiment of the present invention, the language model 131 is an LLM (Large Language Model) that is to say a language model which includes a deep neural network comprising a large number of internal parameters typically on the order of a billion or more and a converter of textual data into tokens and vice versa.

[0090] LLM models are mathematical functions whose input and output are lists of numbers (tokens). In order for them to accept textual data as input and provide texts as output, a conversion is applied to this textual data to obtain tokens (numbers), and a conversion is applied to the tokens to obtain textual data from which the output texts of the LLM model are generated. This conversion is implemented by a bijective function. (called a tokenizer in English) based on a predefined matching (vocabulary). This bijective function is generally first fitted to a complete training text dataset, then frozen, before the deep neural network is trained. Another characteristic of the bijective function is to compress the text data presented as input to the LLM model to save computation time during training and inference of the deep neural network.

[0091] According to a particular and non-limiting embodiment of the present invention, illustrated in [Fig. 3], the LLM model is an autoregressive LLM model based on a transformer-type architecture as described, for example, in the article by Vaswani et al. (Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A., Kaiser, U., & Polosukhin, I. (2017). Attention is All You Need. In Advances in Neural Information Processing Systems. Curran Associates, Inc.). The transformer architecture is a neural network model that relies on self-attention mechanisms to capture the relationships between tokens in a sequence, enabling parallel processing and efficient context learning.

[0092] Basically, an autoregressive LLM based on a transformer-type architecture comprises an encoder (or 'tokenizer') 1311, a decoder (also called a 'tokenizer') 1312, and a deep neural network 1313. The encoder 1311 encodes textual data x, such as words or phrases in natural language, which are present at its input to the LLM model, into a sequence of tokens y, i.e., numbers ('token'), which are present at the input to the deep neural network, based on a predefined mapping (also called a vocabulary) between words and / or phrases in natural language and tokens. The decoder 1312 performs the inverse operation by decoding a sequence of tokens z (output of the deep neural network) to obtain textual data x, on the basis of which a text is generated.The 1313 deep neural network operates on a sequence of tokens y as input (output of the encoder) and generates a new sequence of tokens z as output (input of the decoder).

[0093] The deep neural network 1313 comprises a set of artificial neuron layers.

[0094] For example, each artificial neuron is a perceptron, that is to say a linear classifier comprising generally several inputs and a single output and characterized by an activation function, weights (or synaptic coefficients) and a bias (or threshold).

[0095] For example, a perceptron with n inputs (¾) and a single output o is defined by n weights (Wj, ..., M'„) and a bias (or threshold) 0:

[0096]

[0097]

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104] o = f{z) = 1 SÏ Lr > 0 0 otherwise The output o then results from applying the Heaviside function to the postsynaptic potential z given by: z = L i ^ i x i -o with a nonlinear activation function H(x) given for example by It SJX > 1 The internal parameters of the 1313 neural network are then these weights and biases for the set of perceptrons. The present invention is not limited to this definition of a perceptron, nor to the use of other basic elements forming a layer of the neural network. It is also not limited to the number of perceptrons (or other basic elements) used per layer, nor to the number of layers. The internal parameters of the neural network are usually weights and biases, regardless of the basic elements of the neural network layers. During the training phase of a deep neural network, the network provides a set of output tokens when a set of tokens is presented as input. The input and output token sets correspond to words or sequences of words in natural language. The deep neural network is trained when it provides output token sets that correspond to expected words or sequences of words. More specifically, during the training phase, the internal parameters of the deep neural network in the Language Model 131 are optimized using loss function estimation, which can be of the maximum likelihood type. The Language Model 131 calculates the probability distribution over a given vocabulary for each token in a training sequence presented as input to the deep neural network, and the goal is to maximize the probability that the next 'true' token in the training sequence is predicted. This is achieved by minimizing a cross-entropy loss function defined between the predicted probabilities and the next real token in the training sequence. According to a particular and non-limiting embodiment of the present invention, during the learning phase, the internal parameters of the deep neural network 1313 can be optimized by backpropagation and gradient descent based, for example, on the method described by Robbins and Monro (Robbins, H. and S. Monroe (1951). “A Stochastic Approximation Method.” In The Annals of Mathematical Statistics 22.3, pp. 400-407).

[0105] The learning phase continues until the deep neural network 1313 of the language model 131 reaches a level of performance satisfactory for predicting the next token in a token sequence.

[0106] According to a particular and non-limiting embodiment of the present invention, the language model 131 can undergo several learning phases: a pre-training phase on large sets of public data such as the 'Common Crawl' database and a refinement phase on sets of data dedicated to a specific application, in particular to that of obtaining explanatory information on AD AS functions or automated driving of the vehicle from a question asked by the driver and a context (interior and / or exterior) of the vehicle as explained above.

[0107] Fig. 4 schematically illustrates a device 2 configured to control a sound diffusion system in a vehicle passenger compartment for messages relating to explanatory information on vehicle driver assistance or automated vehicle driving functions, according to a particular and non-limiting embodiment of the present invention.

[0108] Device 2 corresponds for example to a device embedded in the vehicle, for example a computer.

[0109] Device 2 is, for example, configured to implement at least one step of the process described opposite Figures 1, 2, 3, and 5. Examples of such a device 2 include, but are not limited to, embedded electronic equipment such as a vehicle's on-board computer, an electronic control unit such as an ECU (Electronic Control Unit), a smartphone, a tablet, or a laptop computer. The elements of device 2, individually or in combination, may be integrated into a single integrated circuit, into several integrated circuits, and / or into discrete components. Device 2 may be implemented in the form of electronic circuits or software (or computer) modules, or a combination of electronic circuits and software modules.

[0110] The device 2 comprises one (or more) processor(s) 20 configured to execute instructions for carrying out at least one step of the process and / or for executing instructions from the software embedded in the device 2. The processor 20 may include integrated memory, an input / output interface, and various circuits known to those skilled in the art. The device 2 further comprises at least one memory 21, corresponding, for example, to volatile and / or non-volatile memory, and / or includes a memory storage device that may include volatile and / or non-volatile memory, such as EEPROM, ROM, PROM, RAM, DRAM, SRAM, flash, magnetic or optical disk. One or more processors can be used to implement the above computing units.

[0111] The computer code of the embedded software(s) including the instructions to be loaded and executed by the processor is for example stored on memory 21.

[0112] According to various particular and non-limiting embodiments, the device 2 is coupled in communication with other similar devices or systems and / or with communication devices, for example a TCU (Telematic Control Unit), for example via a communication bus or through dedicated input / output ports.

[0113] The deep neural network 1313 of the language model 131 can be trained by a remote device, such as a server or a series of servers, communicating with device 2 via a communication network. The internal parameters of the deep neural network 1313 of the language model 131 can then be transmitted by this remote device to device 2, which receives them. Device 2 is then able to implement the process of [Fig. 5].

[0114] According to a particular and non-limiting embodiment of the present invention, the device 2 comprises a block 22 of interface elements for communicating with external devices, for example, a remote server providing road traffic conditions and / or weather forecasts. The interface elements of the block 22 comprise one or more of the following interfaces: - radio frequency RF interface, for example of the Wi-Fi® type (according to IEEE 802.11), for example in the 2.4 or 5 GHz frequency bands, or of the Bluetooth® type (according to IEEE 802.15.1), in the 2.4 GHz frequency band, or of the Sigfox type using UBN (Ultra Narrow Band) radio technology, or LoRa in the 868 MHz frequency band, LTE (Long-Term Evolution), LTE-Advanced; - USB interface (from the English "Universal Serial Bus" or "Universal Serial Bus" in French); - HDMI interface (from the English "High Definition Multimedia Interface", or "High Definition Multimedia Interface" in French); - LIN interface (from the English "Local Interconnect Network", or in French "Réseau interconnecté local").

[0115] Data, for example internal parameters of the trained deep neural network 1313 or data representative of a driving scenario, a road environment around the vehicle and / or road traffic and environmental conditions, are for example loaded into the device 2 via the interface block 802 using a Wi-Fi® network such as according to IEEE 802.11, an ITS G5 network based on IEEE 802.1 Ip or a mobile network such as a 4G (or 5G) network based on the LTE (Long Term Evolution) standard defined by the 3GPP consortium, including an LTE-V2X network.

[0116] According to another particular and non-limiting embodiment, the device 2 includes a communication interface 23 which enables communication with other devices (such as the system for perceiving the external road environment close to the vehicle, said at least one contextual data capture system, the driver emotion detection system, the voice recognition system) via a communication channel 24. The communication interface 23 corresponds, for example, to a transmitter configured to transmit and receive information and / or data via the communication channel 24.Communication interface 23 corresponds, for example, to a wired network of the CAN type (Controller Area Network), CAN FD (Controller Area Network Flexible Data-Rate), FlexRay (standardized by ISO 17458) or Ethernet (standardized by ISO / IEC 802-3).

[0117] Internal parameters of the trained deep neural network 1313 can be received via the communication interface 23.

[0118] According to a particular and non-limiting embodiment of the present invention, the device 2 can provide output signals to one or more external devices, such as a display screen 25, touch or not, one or more speakers 26 and / or other peripherals 27 via output interfaces 28, 29 and 30 respectively. According to a variant, one or more of the external devices is integrated into the device 2.

[0119] Figure 5 illustrates a flowchart of the different stages of a method for controlling a sound diffusion system in the passenger compartment of a vehicle for messages relating to explanatory information on vehicle driving assistance functions or automated vehicle driving, according to a particular and non-limiting embodiment of the present invention.

[0120] In a step 51, the first data 100 of text type is obtained from the speech recognition system 10.

[0121] In a step 52, the second data 110 of text type is obtained from said at least one contextual data capture system 11.

[0122] In a step 53, the third data 120 of type text is obtained from the text type data fusion system 12 from the first data 100 and the second data 110.

[0123] In a step 54, the third data 120 is submitted to the conversational system 13.

[0124] In a step 55, the fourth data 130 is obtained from the conversational system 13.

[0125] In a step 56, the sound diffusion system 14 is controlled to diffuse the fourth data 130 into the vehicle's passenger compartment.

[0126] According to one variant, the variants and examples described in relation to Figures 1 to 3 apply to the steps of the process in [Fig. 5].

[0127] Of course, the present invention is not limited to the embodiments described above but extends to a method for controlling a sound system in the passenger compartment of a vehicle for displaying messages relating to explanatory information about vehicle driver assistance or automated vehicle driving functions, which would include secondary steps without departing from the scope of the present invention. The same would apply to a device configured for implementing such a method.

[0128] The present invention also relates to a vehicle, for example an automobile or more generally an autonomous land-powered vehicle, comprising device 2 of [Fig.4].

Claims

Demands

1. A method for controlling a sound system in a vehicle's passenger compartment, for a message relating to explanatory information about vehicle driver assistance or automated vehicle driving functions, said method being implemented by at least one processor and comprising the following steps: - obtaining (51), via a speech recognition system, a first data item (100) of text type representing a question asked by a vehicle occupant and relating to explanatory information about one of said vehicle driver assistance or automated vehicle driving functions; - obtaining (52), from at least one contextual data capture system, a second data item (110) of text type relating to a current vehicle driving context;- obtaining (53), from a text-type data fusion system, a third data item (120) of text type from the first data item (100) and the second data item (110), both of text type; - submitting (54) the third data item (120) to a conversational system based on a language model having access to a database comprising data representative of explanatory information on the vehicle's driver assistance or automated driving functions; - obtaining (55), from the conversational system, a fourth data item (130) representative of a response relating to the explanatory information on the vehicle's driver assistance or automated driving functions in the database corresponding to the third data item (120); - controlling (56) the sound system for broadcasting the fourth data item (130) in the vehicle's passenger compartment.

2. A method according to claim 1, wherein the second input comprises at least one word or sequence of words from the natural language representative of a context external to the vehicle, said external context being obtained from a current geographical position of the vehicle.

3. A method according to claim 2, wherein the external context is defined by one of the following data or one of their combinations; - data representative of a driving scenario; - data representative of road traffic and environmental conditions.

4. A method according to any one of the preceding claims, wherein the second data comprises a word or sequence of words from the natural language representative of an interior context of the vehicle, said interior context being obtained from at least one on-board sensor of the vehicle.

5. A method according to claim 4, wherein the internal context is defined by one of the following data or one of their combinations: - data representing a vehicle speed; - data representing a vehicle acceleration; - data representing an activation of one of the AD AS or automated driving functions of the vehicle; - data representing an emotion of the vehicle driver; - data representing the number of occupants of the vehicle; - data representing the number of vehicle occupants conversing; - data representing a road environment around the vehicle.

6. A method according to any one of the preceding claims, wherein the language model comprises a trained deep neural network.

7. Device (2) for controlling a sound diffusion system in a vehicle's passenger compartment, for a message relating to explanatory information on vehicle driving assistance functions or automated vehicle driving, said device (2) comprising a memory (21) associated with at least one processor (20) configured for carrying out the steps of the method according to any one of claims 1 to 6.

8. Computer program comprising instructions for carrying out the method according to any one of claims 1 to 6, when such instructions are executed by a processor.

9. 19 Computer-readable recording medium on which is recorded a computer program comprising instructions for carrying out the steps of the process according to any one of claims 1 to 6.

10. Vehicle comprising a device according to claim 7.