CN115240677A - Voice interaction method, device and equipment for vehicle cabin - Google Patents

Abstract

There is provided a voice interaction method for a vehicle cabin comprising a plurality of zones respectively corresponding to a plurality of vehicle seats, the plurality of zones being collectively configured with a current conversation cache for storing current rounds of conversations, and each zone of the plurality of zones being configured with a respective zone cache for storing historical semantic content from the zone within a preset time window, the method comprising: obtaining current semantic content from a first one of the plurality of sound zones; updating the current multi-turn conversation based on the current conversation cache, the sound zone cache of the first sound zone and the sound zone caches of the rest sound zones except the first sound zone in the plurality of sound zones, wherein the updated current multi-turn conversation is associated with the current semantic content; and processing current semantic content associated with the updated current plurality of sessions.

Description

Voice interaction method, apparatus and device for vehicle cockpit

technical field

The present disclosure relates to the field of vehicles, and in particular to a voice interaction method for a vehicle cabin, a voice interaction device for a vehicle cabin, a computer device for a vehicle cabin, a vehicle including the above-mentioned voice interaction device or computer device, and a storage device. media and computer program products.

Background technique

In recent years, with the rapid development of artificial intelligence technology, voice interaction technology also has a practical level. With the continuous increase of the number of private cars, the vehicle-mounted voice interaction system has become the standard configuration of the car-machine system. At the same time, people are more and more eager to communicate with the vehicle-machine system naturally and conveniently, so as to improve the interactive experience and driving safety. Therefore, further improving the vehicle voice interaction system is one of the important tasks to realize the intelligentization of vehicle cockpit voice interaction.

The approaches described in this section are not necessarily approaches that have been previously conceived or employed. Unless otherwise indicated, it should not be assumed that any of the approaches described in this section qualify as prior art merely by virtue of their inclusion in this section. Similarly, unless otherwise indicated, the issues raised in this section should not be considered to be recognized in any prior art.

SUMMARY OF THE INVENTION

Embodiments of the present disclosure provide a voice interaction method for a vehicle cabin, a voice interaction device for a vehicle cabin, a computer device, a vehicle including the above-mentioned voice interaction device or voice interaction device, a storage medium, and a computer program product.

According to an aspect of the present disclosure, there is provided a voice interaction method for a vehicle cabin, the vehicle cabin including a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones being jointly configured with a current session cache to For storing the current multi-round conversation, and each sound region in the plurality of sound regions is configured with a corresponding sound region buffer for storing historical semantic content from the sound region within a preset time window, the method includes: obtaining data from The current semantic content of the first sound region in the plurality of sound regions; based on the current session buffer, the sound region buffer of the first sound region and the remaining sound regions of the plurality of sound regions except the first sound region The sound area cache updates the current multi-round session, wherein the updated current multi-round session is associated with the current semantic content; and processes the current semantic content associated with the updated current multi-round session.

According to another aspect of the present disclosure, there is provided a voice interaction device for a vehicle cabin, the vehicle cabin including a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones being jointly configured with a current session cache For storing the current multi-round conversation, and each sound region in the plurality of sound regions is configured with a corresponding sound region buffer for storing historical semantic content from the sound region within a preset time window, the device comprises: obtaining module, the obtaining module is configured to obtain the current semantic content from the first sound region in the plurality of sound regions; the updating module is configured to be based on the current session cache, the sound region cache of the first sound region and the sound region buffers of the remaining sound regions in the plurality of sound regions except the first sound region, updating the current multi-round conversation, wherein the updated current multi-round conversation is associated with the current semantic content; and a processing module, the The processing module is configured to process the current semantic content associated with the updated current multi-round session.

According to yet another aspect of the present disclosure, there is provided a computer device for a vehicle cabin, the vehicle cabin including a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones being jointly configured with a current session cache to For storing the current multi-round conversation, and each sound region in the plurality of sound regions is configured with a corresponding sound region buffer for storing historical semantic content from the sound region within a preset time window, the computer equipment includes: at least a processor; and at least one memory having stored thereon a computer program that, when executed by the at least one processor, causes the at least one processor to implement the above-described method.

According to yet another aspect of the present disclosure, there is provided a vehicle comprising the above-mentioned voice interaction device or computer equipment.

According to yet another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing a computer program, the computer program comprising instructions that, when executed by a processor, cause the processor to perform the above-described method.

According to yet another aspect of the present disclosure, there is provided a computer program product comprising instructions which, when executed by a processor, cause the processor to perform the above-described method.

According to the embodiments of the present disclosure, it is possible to realize the simultaneous interaction of multi-sound zones and multi-users in the vehicle cockpit with the vehicle machine, and to link different user sessions with contextual relevance to avoid the intention of mapping different users' voices into separate ones. , thereby improving the accuracy of the voice interaction in the vehicle cockpit and helping to improve the user's car experience.

These and other aspects of the present disclosure will be apparent from and elucidated with reference to the embodiments described hereinafter.

Description of drawings

Further details, features and advantages of the present disclosure are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings. The accompanying drawings illustrate the embodiments by way of example and constitute a part of the specification, and together with the written description of the specification serve to explain exemplary implementations of the embodiments. The shown embodiments are for illustrative purposes only and do not limit the scope of the claims. Throughout the drawings, the same reference numbers refer to similar but not necessarily identical elements. In the attached image:

1 is a schematic diagram illustrating an example system in which various methods described herein may be implemented, according to an example embodiment;

FIG. 2 is a flowchart illustrating a voice interaction method for a vehicle cockpit according to an exemplary embodiment;

3 is a flowchart illustrating a method of updating a current multi-round session according to an exemplary embodiment;

4 is a logical block diagram illustrating a voice interaction process for a vehicle cabin in accordance with an exemplary embodiment;

5 is a block diagram illustrating a voice interaction device for a vehicle cabin according to an exemplary embodiment; and

FIG. 6 is a block diagram illustrating an example computer device that can be applied to the example embodiments.

Detailed ways

In the present disclosure, unless otherwise specified, the use of the terms "first", "second", etc. to describe various elements is not intended to limit the positional relationship, timing relationship or importance relationship of these elements, and such terms are only used for Distinguish one element from another. In some examples, the first element and the second element may refer to the same instance of the element, while in some cases they may refer to different instances based on the context of the description.

The terminology used in the description of the various described examples in this disclosure is for the purpose of describing particular examples only and is not intended to be limiting. Unless the context clearly dictates otherwise, if the number of an element is not expressly limited, the element may be one or more. As used herein, the term "plurality" means two or more, and the term "based on" should be construed as "based at least in part on." Furthermore, the terms "and/or" and "at least one of" encompass any and all possible combinations of the listed items.

Before introducing exemplary embodiments of the present disclosure, several terms used herein are first explained.

As used herein, the term "sound zone" refers to a spatial division that is equipped with a microphone or array of microphones that capture user speech and/or ambient sound, and is associated with a seating position in a vehicle cabin. The vehicle cockpit space can be divided into sound zones according to actual needs. For example, the vehicle cockpit space can be divided into two sound zones according to the front and rear seats (the actual situation includes but not limited to the vehicle cockpit space with two rows of seats, such as , when the vehicle cockpit has more than two rows of seats, the vehicle cockpit space can be divided into more than two sound areas accordingly), and the vehicle cockpit space can be divided into corresponding different sound areas according to the driver's seat, co-pilot seat and passenger seat, and many more.

As used herein, the term "sound zone buffer" refers to a memory device corresponding to a sound zone that stores user speech and/or ambient sound or processed signals thereof captured by a microphone or microphone array equipped with the sound zone. In the case of no explicit description, when referring to the voice zone cache, it can refer to the objects stored in the voice zone cache, for example, including but not limited to the natural language understanding (NLU) from the voice zone. Processed semantic content, etc.

As used herein, the term "current session cache" refers to a storage associated with a session(s) currently in progress by a user throughout the vehicle cabin space, storing microphones equipped by the sound zone(s) involved in the session or A storage device for user speech and/or ambient sound or its processed signals collected by the microphone array. As an example, in the actual signal processing process, the objects stored in the current session cache may include the user's voice and/or the ambient sound or the processed signals thereof collected by the microphones or microphone arrays of the individual sound zones that make up the current session. Cascading in chronological order. Unless otherwise specified, when referring to the current session cache, it may refer to objects stored in the current session cache, for example, including but not limited to the current round sessions, etc.

As used herein, the term "semantic content" refers to the original speech signal after being processed by natural language understanding (NLU), containing contextual information (eg, whether there is a continuation with other speech), session state (eg, whether it is a paragraph or not) the beginning or end of a conversation) and intent (eg, whether to include a query for specific information), a processed signal with rich details.

As used herein, the term "multi-turn session" refers to a session that includes more than one speech signal that makes up the session, which is usually completed by multiple users, but does not preclude the possibility of a single user having multiple sessions, for example, when a single After the user speaks a voice with a query purpose, he then speaks another voice to supplement the aforementioned query.

From the traditional one-question and one-answer voice interaction system to the more popular multi-round question-and-answer voice interaction system, the way of human-computer interaction tends to be close to human-human interaction. Voice interaction usually involves speech recognition, natural language understanding, and speech synthesis, among others. In smart cockpits, speakers are often not unique. In this regard, in the related art, there is an in-vehicle voice interaction system that can realize simultaneous interaction of multiple passengers. These in-vehicle voice interaction systems are realized by creating multiple voice interaction links for multiple sound zones in the cockpit (that is, dividing the cockpit into multiple sound zones (seats), and collecting voice signals in different areas through a microphone array) of. When one or more voice interaction links monitor the voice signal in the sound zone, the vehicle terminal switches the one or more voice interaction links to the voice processing state, and the voice processing state is used to process the passengers in the corresponding voice zone. input voice signal. Although this method supports switching of conversation objects, unfortunately, only one voice interaction link is in the speech processing state at the same time, so this method is still a serial single-session interaction. In addition, in the related art, there is also a method of sending the signals of multiple sound zones to a preset cloud server for semantic recognition so as to generate corresponding voice commands to realize in-vehicle voice interaction, but such methods cannot determine the Whether there is a contextual relationship between speech signals.

In order to solve the above technical problems, according to one or more embodiments of the present disclosure, a new voice interaction method for a vehicle cockpit is proposed. The method updates the current multi-round conversation based on the current session buffer, the sound area buffer of the first sound area, and the sound area buffers of the other sound areas in the plurality of sound areas except the first sound area, so that when The current semantic content associated with the updated current multi-round session is better handled with contextual relevance taken into account. Through the above method, the simultaneous interaction between multi-sound zones and multi-users in the vehicle cockpit and the vehicle machine can be realized, and different user sessions with contextual relevance can be linked to avoid mapping the voices of different users into separate intentions, thereby improving the The accuracy of the voice interaction in the vehicle cockpit can help improve the user's car experience. Exemplary embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.

FIG. 1 is a schematic diagram illustrating an example system 100 in which various methods described herein may be implemented, according to an example embodiment.

Referring to FIG. 1 , the system 100 includes an in-vehicle system 110 , a server 120 , and a network 130 communicatively coupling the in-vehicle system 110 and the server 120 .

The in-vehicle system 110 includes a display 114 and an application program (APP) 112 that can be displayed via the display 114 . The application 112 may be an application installed by default on the in-vehicle system 110 or downloaded and installed by the user 102, or as an applet of a lightweight application. In the case where the application 112 is an applet, the user 102 can search for the application 112 in the host application (eg, by the name of the application 112, etc.) or scan the graphic code (eg, barcode, QR code, etc.) of the application 112 etc.), etc., the application program 112 is directly run on the in-vehicle system 110 without installing the application program 112 . In some embodiments, in-vehicle system 110 may include one or more processors and one or more memories (not shown), and in-vehicle system 110 is implemented as an in-vehicle computer. In some embodiments, in-vehicle system 110 may include more or less display screen 114 (eg, no display screen 114), and/or one or more speakers or other human-computer interaction devices. In some embodiments, in-vehicle system 110 may not communicate with server 120 .

Server 120 may represent a single server, a cluster of multiple servers, a distributed system, or a cloud server that provides basic cloud services such as cloud database, cloud computing, cloud storage, cloud communication. It will be appreciated that although the server 120 is shown in FIG. 1 in communication with only one in-vehicle system 110, the server 120 may provide background services for multiple in-vehicle systems simultaneously.

The network 130 allows wireless communication and information exchange between vehicle-X ("X" means vehicle, road, pedestrian, or the Internet, etc.) according to agreed communication protocols and data exchange standards. Examples of network 130 include a local area network (LAN), a wide area network (WAN), a personal area network (PAN), and/or a combination of communication networks such as the Internet. Network 130 may be a wired or wireless network. In one example, the network 130 may be an in-vehicle network, an inter-vehicle network, and/or an in-vehicle mobile Internet.

For the purpose of the embodiments of the present disclosure, in the example of FIG. 1 , the application 112 may be an electronic map application, which may provide various functions based on the electronic map, such as navigation, route query, location search, etc. Wait. Correspondingly, the server 120 may be a server used with an electronic map application. The server 120 can provide online map services, such as online navigation, online route query, and online location search, to the application 112 running in the vehicle system 110 based on the road network data. Alternatively, the server 120 may also provide the road network data to the in-vehicle system 110, and the application 112 running in the in-vehicle system 110 provides a local map service according to the road network data.

FIG. 2 is a flowchart illustrating a voice interaction method 200 for a vehicle cabin, according to an exemplary embodiment. The method 200 may be performed at an in-vehicle system (eg, the in-vehicle system 110 shown in FIG. 1 ), that is, the execution subject of each step of the method 200 may be the in-vehicle system 110 shown in FIG. 1 . In some embodiments, method 200 may be performed at a server (eg, server 120 shown in FIG. 1 ). In some embodiments, method 200 may be performed by a combination of an in-vehicle system (eg, in-vehicle system 110 ) and a server (eg, server 120 ). Hereinafter, each step of the method 200 will be described by taking the execution subject as the in-vehicle system 110 as an example. Here, the vehicle cockpit in which the in-vehicle system 110 is located includes a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones are jointly configured with a current session buffer for storing the current multi-round conversation, and the plurality of sound zones are Each sound area in is configured with a corresponding sound area buffer for storing historical semantic content from the sound area within a preset time window.

As shown in FIG. 2, method 200 includes:

Step S210, obtains the current semantic content from the first sound section in the plurality of sound sections;

Step S220, based on the current session cache, the voice zone cache of the first voice zone, and the voice zone caches of the rest of the multiple voice zones except the first voice zone, update the current multi-round session, wherein the updated current Multiple rounds of conversations are associated with the current semantic content; and

Step S230, processing the current semantic content associated with the updated current multi-round session.

The various steps of method 200 are described in detail below.

In step S210, the current semantic content from the first sound region of the plurality of sound regions is acquired. The first sound zone may be any one of a plurality of sound zones included in the vehicle cabin and corresponding to a plurality of vehicle seats respectively. For example, the first sound zone may be the sound zone corresponding to the driver's seat or may be the sound zone corresponding to the front passenger seat and/or the rear passenger seat, which is not limited in the present disclosure. The current semantic content may be a processed signal obtained after the current original speech signal from the first sound region is processed by natural language understanding NLU. In some examples, the current raw speech signal from the first sound zone may be collected by a microphone or microphone array equipped with the first sound zone.

In step S220, the current multi-round session is updated based on the current session buffer, the sound area buffer of the first sound area, and the sound area buffers of the remaining sound areas of the plurality of sound areas except the first sound area. Further, the updated current multi-round conversation is associated with the current semantic content. The content stored in the current session cache is associated with the current session (multiple) users in the entire vehicle cabin space, and the content stored in the voice zone cache of the first sound zone is related to the current conversation of the user in the first voice zone. The voices or the user's historical voices are associated, and the content stored in the voice zone caches of the remaining voice zones in the multiple voice zones except the first voice zone is associated with the user's currently spoken voice or the user's historical voice in the corresponding voice zone . In this way, the problem of NLU cross-linking can be avoided (that is, the intentions included in multiple sessions within a multi-tone area may be merged and understood as the same intention by NLU), and different user sessions with contextual relevance can be linked to avoid Simple mapping of speech from different users into separate intents.

In step S230, the current semantic content associated with the updated current multi-round session is processed. Since the current multi-round session has been updated in step S220, the current semantic content associated with the updated current multi-round session can reflect the contextual information of the current multi-round session in which it is located, so the vehicle and the machine can be more targeted. It can give a response to the current semantic content, thereby improving the accuracy of human-machine voice interaction in the vehicle cockpit.

According to the embodiments of the present disclosure, the above-mentioned method 200 overcomes the deficiencies in the related art that the in-vehicle voice interaction system responds to multi-voice input in a serial single-session interaction, but cannot truly respond to multi-speech area and multi-session intentions in parallel, and cannot judge multiple Whether there is a defect in the contextual relationship between the speech signals of the vocal area. The method 200 updates the current multi-round session based on the current session cache, the voice zone cache of the first voice zone, and the voice zone buffers of other voice zones in the plurality of voice zones except the first voice zone. Taking relevancy into account can better handle the current semantic content associated with the updated current multi-round session, thereby better understanding and responding to the user's intent and improving the user's car experience.

FIG. 3 is a flowchart illustrating a method 300 of updating a current multi-round session, according to an exemplary embodiment. The method 300 may be performed at an in-vehicle system (eg, the in-vehicle system 110 shown in FIG. 1 ), that is, the execution subject of each step of the method 300 may be the in-vehicle system 110 shown in FIG. 1 . In some embodiments, method 300 may be performed at a server (eg, server 120 shown in FIG. 1 ). In some embodiments, method 300 may be performed by a combination of an in-vehicle system (eg, in-vehicle system 110 ) and a server (eg, server 120 ). Hereinafter, each step of the method 300 will be described by taking the execution subject as the in-vehicle system 110 as an example. Here, the vehicle cockpit in which the in-vehicle system 110 is located includes a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones are jointly configured with a current session buffer for storing the current multi-round conversation, and the plurality of sound zones are Each sound area in is configured with a corresponding sound area buffer for storing historical semantic content from the sound area within a preset time window.

As shown in FIG. 3, method 300 includes:

Step S310, determine whether there is historical semantic content stored in the sound region cache of the first sound region;

Step S320, determining whether the current semantic content and the historical semantic content belong to the same multi-round conversation in response to determining that the historical semantic content is stored in the sound region cache of the first sound region; and

Step S330, in response to determining that the current semantic content and the historical semantic content belong to the same multi-round session, store the current semantic content and the historical semantic content together in the current session cache to update the current multi-round session.

The various steps of the method 300 are described in detail below with reference to FIG. 4 .

FIG. 4 is a logical block diagram illustrating a voice interaction process 400 for a vehicle cabin, according to an exemplary embodiment. The workflow of process 400 may include acquiring a new sound source signal from one of the multi-tone regions through a microphone or microphone array, processing the sound source signal, using Automatic Speech Recognition (ASR) and/or Or NLU completes the semantic content analysis of the sound source signal. In a set of multi-round sessions, when multiple sound source signals (which may be accompanied by an intentional understanding request) have been received, the ASR and/or NLU combine the cached state of that voice zone, the cached state of the remaining voice zones, and the vehicle cockpit The corresponding current session cache is used to determine whether the newly received sound source signal and one or more of the multiple received sound source signals belong to the same multi-round session, and the intention understanding at the moment is given accordingly, and Process the current request corresponding to the newly received sound source signal (for example, new semantic content), and update the current session buffer and sound area buffer. If none of them belong to the multi-round dialogue, the newly received sound source signal is newly created as the current multi-round session in the current session buffer (for example, to replace the content stored in the current session buffer) and the sound area buffer is updated.

The various blocks of process 400 are described in detail below.

At block 401, new semantic content is awaited to be received. The semantic content may be a processed signal from the current raw speech signal from the phonetic region after being processed by natural language understanding NLU. In some examples, the current raw speech signal from the sound zone may be collected by a microphone or microphone array with which the sound zone is equipped.

At block 402, new semantic content for region i is received. The sound zone i may be any one of a plurality of sound zones included in the vehicle cabin and corresponding to a plurality of vehicle seats respectively. For example, the sound zone i may be the sound zone corresponding to the driver's seat or the sound zone corresponding to the front passenger seat and/or the rear passenger seat, which is not limited in the present disclosure.

At block 403, it is determined whether historical semantic content is stored in the buffer of sound zone i. Block 403 may correspond to step S310 of FIG. 3 : determining whether historical semantic content is stored in the sound region buffer of the first sound region, and the sound region i corresponds to the first sound region at this time. Historical semantic content can refer to historical speech signals obtained after being processed by natural language understanding NLU, containing contextual information (for example, whether there is a continuation with other speech), conversation state (for example, whether it is the beginning or end of a dialogue) and A processed signal that is rich in detail such as intent (eg, whether to include a query for specific information).

According to an embodiment of the present disclosure, the query time window length of the tone zone cache can be adjusted by custom.

When the determination of block 403 is yes, process 400 proceeds to block 404 .

At block 404, it is determined whether the new semantic content belongs to the same multi-round session as the historical semantic content in the cache of zone i. Block 404 may correspond to step S320 of FIG. 3 : determining whether the current semantic content and the historical semantic content belong to the same multi-round conversation in response to determining that the historical semantic content is stored in the sound region cache of the first sound region. It should be noted that the method for judging whether two semantic contents belong to the same multi-round conversation may include ASR, NLU and other speech processing technologies known in the art.

When the judgment result of block 404 is yes (that is, the semantic content newly received in sound area i is contextually related to the historical semantic content previously received in this sound area i, so that the new semantic content in sound area i is The process 400 proceeds to block 405 when the content and the historical semantic content constitute the same multi-round conversation).

At block 405, the current session cache is updated. Specifically, the newly received semantic content is stored in the current session cache together with its associated historical semantic content to overwrite the content stored in the current session cache. Block 405 may correspond to step S330 of FIG. 3: in response to determining that the current semantic content and the historical semantic content belong to the same multi-round session, store the current semantic content together with the historical semantic content in the current session cache to update the current multi-round session. Similar to the query time window of the sound zone cache, a customizable query time window length can also be set for the current session cache. It should be noted here that, in the case that the query time window of the current session cache is not long enough to cover both the current semantic content and the historical semantic content to be overwritten in the current session cache, the time difference of the historical semantic content is relatively high. The early part may not be stored in the current session cache. That is, the content stored in the current session cache is the closest to the current time point. At this point, method 300 is complete.

Other exemplary embodiments of the present disclosure are described with continued reference to FIG. 4 .

Continuing with the above-described exemplary embodiment, when process 400 proceeds via block 403 to block 404 and the decision at block 404 is no (that is, the semantic content newly received in phoneme i is the same as the process 400 may proceed from block 404 to block 400 when the received historical semantic content does not have contextual relevance such that the new semantic content in voice zone i and the historical semantic content in voice zone i cannot constitute the same multi-round conversation) 407.

At block 407, it is determined whether the new semantic content and the current multi-round session in the current session cache belong to the same multi-round session. The content stored in the current session cache is associated with the user's current session(s) throughout the vehicle cabin space. For example, if the new semantic content is "to have a parking space", and the current multi-round sessions stored in the current session cache include "buy luggage" and "nearest shopping mall", it can be determined that the new semantic content is related to the current session cache. The stored current multi-round session belongs to the same multi-round session, where the intention of the multi-round session can be understood as going to the nearest shopping mall with parking spaces to buy bags, and the content to be overwritten in the current session cache is "To be Parking available", "Buy luggage" and "Nearest mall". Thus, process 400 may proceed from block 407 to block 405 when the determination of block 407 is yes.

According to an embodiment of the present disclosure, the above method may optionally include an additional step of: in response to determining that the current semantic content and the historical semantic content do not belong to the same multi-round session, determining whether the current semantic content and the current multi-round session belong to the same multi-round session a round session; and in response to determining that the current semantic content and the current multi-round session belong to the same multi-round session, storing the current semantic content with the current multi-round session in the current session cache to update the current multi-round session.

Continuing with the exemplary embodiment described above, when process 400 proceeds to block 407 via blocks 403 , 404 and the determination at block 407 is negative, process 400 may proceed from block 407 to block 408 . At block 408, it is determined whether the new semantic content and the historical semantic content in the buffers of the remaining sound regions j of the plurality of sound regions except sound region i belong to the same multi-round conversation. In some cases, the new semantic content currently received in voice zone i may be associated neither with the historical semantic content of voice zone i nor with the current multi-round session stored in the current session cache. For example, the new semantic content of sound area i is "limited line end number", the historical semantic content of sound area i is "how many degrees is the temperature today", and the current multi-round sessions stored in the current session cache include "buy luggage", " The nearest shopping mall", then neither the correlation between the new semantic content currently received in sound area i and the historical semantic content of sound area i, nor the new semantic content currently received in sound area i can be determined. Correlation with the current multi-round session stored in the current session cache. For this reason, it is necessary to judge whether the new semantic content in sound region i and the historical semantic content in the buffers of other sound regions j in multiple sound regions except for sound region i belong to the same multi-round conversation.

When the new semantic content in sound region i and the historical semantic content in the buffers of other sound regions j in multiple sound regions except sound region i belong to the same multi-round session (for example, the historical semantic content in the cache of sound region j When the content is "the day of the week tomorrow", so that the intention of the multi-round conversation formed by the two can be understood as tomorrow's road has unlimited lines, what is the limit of the end number of the line), the process 400 can proceed from block 408 to block 405, in block 408 In 405, the new semantic content in the sound zone i and the historical semantic content in the buffer of the sound zone j are stored in the current session cache, so as to update the content stored in the current session cache.

According to an embodiment of the present disclosure, the above method may optionally include an additional step of: in response to determining that the current semantic content and the current multi-round conversation do not belong to the same multi-round conversation, determining that the current semantic content and the plurality of voice regions except the first Whether the historical semantic content stored in the sound region cache of the remaining sound regions other than the sound region belongs to the same multi-round session; The stored historical semantic content belongs to the same multi-round session, and the current semantic content is stored in the current session cache together with the historical semantic content stored in the voice zone cache of the second voice zone to update the current multi-round session. Here, the second sound zone is a sound zone different from the first sound zone.

Continuing with the exemplary embodiment described above, when process 400 proceeds to block 408 via blocks 403 , 404 , 407 and the determination at block 408 is negative, process 400 may proceed from block 408 to block 409 . When the new semantic content in sound area i does not belong to the same multi-round session as the historical semantic content in the buffers of the remaining sound area j in multiple sound areas except sound area j (in addition, as mentioned above, the sound process 400 may proceed from block 408 to Block 409, in block 409, replace the current multi-round session in the current session cache with the new semantic content in zone i.

According to an embodiment of the present disclosure, the above method may optionally include an additional step of: in response to determining the current semantic content and the historical semantics stored in the voice zone caches of the remaining voice zones of the plurality of voice zones except the first voice zone If the content does not belong to the same multi-round session, store the current semantic content in the current session cache to update the current multi-round session.

Returning to block 403 , when the determination of block 403 is negative, process 400 may proceed from block 403 to block 407 . Further, when the determination of block 407 is positive, process 400 may then proceed from block 407 to block 405 . This indicates that the new semantic content received in the sound area i may be the initial semantic content of the sound area (that is, the sound area buffer of the sound area i was empty before this), and it is different from the current number stored in the current session cache. Round conversations (eg, which may consist of semantic content from the rest of the phonetic regions other than phoneme i) are associated.

According to an embodiment of the present disclosure, the above method may optionally include an additional step of determining whether the current semantic content and the current multi-round conversation belong to the same multi-session, in response to determining that no historical semantic content is stored in the voice zone cache of the first voice zone. a round session; and in response to determining that the current semantic content and the current multi-round session belong to the same multi-round session, storing the current semantic content with the current multi-round session in the current session cache to update the current multi-round session.

Continuing with the exemplary embodiment described above, process 400 may proceed from block 408 to block 405 when process 400 proceeds to block 408 via blocks 403 , 407 and the determination at block 408 is yes. For example, the situation may be: the new semantic content of sound area i is "limited line end number", there is no historical semantic content in the cache of sound area i (that is, the cache of sound area i is empty), the current session cache stored in the cache The current multi-round conversation includes "buying luggage" and "nearest shopping mall", while the historical semantic content stored in the cache of other sound zones j is "tomorrow's day of the week". Therefore, the new semantic content of sound area i can form the same multi-round conversation with the historical semantic content stored in the cache of the remaining sound area j.

According to an embodiment of the present disclosure, the above method may optionally include an additional step of: in response to determining that the current semantic content and the current multi-round conversation do not belong to the same multi-round conversation, determining that the current semantic content and the plurality of voice regions except the first Whether the historical semantic content stored in the sound region cache of the remaining sound regions other than the sound region belongs to the same multi-round session; The stored historical semantic content belongs to the same multi-round session, and the current semantic content is stored in the current session cache together with the historical semantic content stored in the voice zone cache of the second voice zone to update the current multi-round session. Here, the second sound zone is a sound zone different from the first sound zone.

Continuing with the exemplary embodiment described above, when process 400 proceeds to block 408 via blocks 403 , 407 and the determination at block 408 is negative, process 400 may proceed from block 408 to block 405 . For example, the situation may be: the new semantic content of sound area i is "limited line end number", there is no historical semantic content in the cache of sound area i (that is, the cache of sound area i is empty), the current session cache stored in the cache The current multi-round conversation includes "buying luggage" and "nearest shopping mall", while the historical semantic content stored in the cache of other sound zones j is either empty or has no contextual relevance with "limited line ending number".

According to an embodiment of the present disclosure, the above method may optionally include an additional step of: in response to determining the current semantic content and the historical semantics stored in the voice zone caches of the remaining voice zones of the plurality of voice zones except the first voice zone If the content does not belong to the same multi-round session, store the current semantic content in the current session cache to update the current multi-round session.

Process 400 also includes block 410, wherein the cache of zone i is updated.

According to an embodiment of the present disclosure, the above-mentioned method may optionally include an additional step of: storing the current semantic content in the sound region cache of the first sound region to update the historical semantics stored in the sound region cache of the first sound region content.

Note that process 400 may be performed in a loop. For example, process 400 may re-enter block 401 after proceeding to block 410 in anticipation of receiving new semantic content.

Although various operations are depicted in the figures as being in a particular order, this should not be construed as a requirement that these operations be performed in the particular order shown or in a sequential order action to obtain the desired result.

FIG. 5 is a schematic block diagram illustrating a voice interaction apparatus 500 for a vehicle cabin according to an exemplary embodiment. The vehicle cockpit includes a plurality of sound zones respectively corresponding to the plurality of vehicle seats, the plurality of sound zones are jointly configured with a current session buffer for storing the current multi-round session, and each sound zone of the plurality of sound zones is configured with a corresponding The sound area buffer is used to store the historical semantic content from the sound area within the preset time window.

The apparatus 500 includes: an obtaining module 510 configured to obtain current semantic content from a first sound region of the plurality of sound regions; an updating module 520 configured to be based on the current session cache, the sound region cache of the first sound region and the multiple sound regions. The sound region caches of the remaining sound regions in the sound regions except the first sound region, update the current multi-round conversation, wherein the updated current multi-round conversation is associated with the current semantic content; and a processing module 530, configured to process The current semantic content associated with the updated current multi-round session.

It should be understood that various modules of the apparatus 500 shown in FIG. 5 may correspond to various steps in the method 200 described with reference to FIG. 2 . Thus, the operations, features, and advantages described above with respect to method 200 are equally applicable to apparatus 500 and the modules it includes.

According to the embodiments of the present disclosure, the above-mentioned apparatus 500 overcomes the deficiencies in the related art that the in-vehicle voice interaction system responds to multi-voice input in a serial single-session interaction mode and cannot truly respond to multi-speech area and multi-session intentions in parallel, and cannot determine multiple Whether there is a defect in the contextual relationship between the speech signals of the vocal area. The apparatus 500 updates the current multi-round conversation based on the current session buffer, the sound area buffer of the first sound area, and the sound area buffers of the other sound areas in the plurality of sound areas except the first sound area. Taking relevancy into account can better handle the current semantic content associated with the updated current multi-round session, thereby better understanding and responding to the user's intent and improving the user's car experience.

Although specific functionality is discussed above with reference to specific modules, it should be noted that the functionality of the various modules discussed herein may be divided into multiple modules, and/or at least some of the functionality of multiple modules may be combined into a single module. Performance of an action by a particular module discussed herein includes the particular module performing the action itself, or alternatively the particular module calling or otherwise accessing another component or module that performs the action (or performs the action in conjunction with the particular module). Thus, a particular module that performs an action may include the particular module that performs the action itself and/or another module that the particular module calls or otherwise accesses that performs the action. As used herein, the phrase "perform action Z based on A, B, and C" may mean based on A only, based on B only, based on C only, based on A and B, based on A and C, based on B and C, or based on A and B and C to perform action Z.

It should also be understood that various techniques may be described herein in the general context of software hardware elements or program modules. The various modules described above with respect to FIG. 5 may be implemented in hardware or in hardware in conjunction with software and/or firmware. For example, these modules may be implemented as computer program code/instructions configured to be executed in one or more processors and stored in a computer-readable storage medium. Alternatively, these modules may be implemented as hardware logic/circuitry. For example, in some embodiments, one or more of the acquisition module 510, the update module 520, and the processing module 530 may be implemented together in a System on Chip (SoC). An SoC may include an integrated circuit chip (which includes a processor (eg, a central processing unit (CPU), microcontroller, microprocessor, digital signal processor (DSP), etc.), memory, a or more communication interfaces, and/or one or more components of other circuits), and may optionally execute the received program code and/or include embedded firmware to perform functions.

According to one aspect of the present disclosure, there is provided a computer device for a vehicle cabin, the vehicle cabin including a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones being collectively configured with a current session buffer for A current multi-round session is stored, and each of the plurality of sound regions is configured with a corresponding sound region buffer for storing historical semantic content from the sound region within a preset time window. The computer device includes at least one memory, at least one processor, and a computer program stored on the at least one memory. The at least one processor is configured to execute a computer program to implement the steps of any of the method embodiments described above.

According to an aspect of the present disclosure, there is provided a vehicle comprising the voice interaction apparatus or computer equipment as described above.

According to an aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium having stored thereon a computer program that, when executed by a processor, implements the steps of any of the above-described method embodiments.

According to one aspect of the present disclosure, there is provided a computer program product comprising a computer program that, when executed by a processor, implements the steps of any of the method embodiments described above.

In the following, illustrative examples of such computer devices, non-transitory computer-readable storage media, and computer program products are described in conjunction with FIG. 6 .

FIG. 6 shows an example configuration of a computer device 600 that may be used to implement the methods described herein. For example, server 120 and/or in-vehicle system 110 shown in FIG. 1 may include an architecture similar to computer device 600 . The above-mentioned apparatus 500 or computer equipment may also be implemented in whole or at least in part by computer equipment 600 or similar equipment or systems.

Computer device 600 may include at least one processor 602, memory 604, communication interface(s) 606, display device 608, other input/output (I/O) devices capable of communicating with each other, such as through a system bus 614 or other suitable connection 610 and one or more mass storage devices 612.

The processor 602 may be a single processing unit or multiple processing units, all of which may include single or multiple computing units or multiple cores. Processor 602 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuits, and/or any device that manipulates signals based on operational instructions. Among other capabilities, processor 602 may be configured to retrieve and execute computer-readable instructions, such as program code of operating system 616 , application programs 618 , stored in memory 604 , mass storage device 612 , or other computer-readable media , the program code of other programs 620, and the like.

Memory 604 and mass storage device 612 are examples of computer-readable storage media for storing instructions that are executed by processor 602 to implement the various functions described above. For example, memory 604 may generally include both volatile and non-volatile memory (eg, RAM, ROM, etc.). Additionally, mass storage devices 612 may generally include hard drives, solid state drives, removable media, including external and removable drives, memory cards, flash memory, floppy disks, optical disks (eg, CDs, DVDs), storage arrays, network attached storage , storage area network, etc. Both memory 604 and mass storage device 612 may be collectively referred to herein as memory or computer-readable storage media, and may be non-transitory media capable of storing computer-readable, processor-executable program instructions as computer program code, The computer program code may be executed by processor 602 as a specific machine configured to implement the operations and functions described in the examples herein.

A number of programs may be stored on mass storage device 612 . These programs include operating system 616, one or more application programs 618, other programs 620, and program data 622, and may be loaded into memory 604 for execution. Examples of such applications or program modules may include, for example, computer program logic (eg, computer program code or instructions) for implementing the following components/functions: method 200, method 300 and optional additional steps thereof, apparatus 500, and /or additional embodiments described herein.

Although illustrated in FIG. 6 as being stored in memory 604 of computer device 600 , modules 616 , 618 , 620 and 622 , or portions thereof, may be implemented using any form of computer-readable medium accessible by computer device 600 . As used herein, "computer-readable media" includes at least two types of computer-readable media, namely, computer-readable storage media and communication media.

Computer readable storage media includes volatile and nonvolatile, removable and non-removable media implemented by any method or technology for storage of information such as computer readable instructions, data structures, programs modules or other data. Computer-readable storage media include, without limitation, RAM, ROM, EEPROM, flash memory, or other memory technologies, CD-ROMs, digital versatile disks (DVDs), or other optical storage devices, magnetic cartridges, tapes, magnetic disk storage devices, or other magnetic storage devices device, or any other non-transport medium that can be used to store information for access by a computer device. In contrast, communication media may embody computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism. Computer-readable storage media, as defined herein, do not include communication media.

One or more communication interfaces 606 are used to exchange data with other devices, such as through a network, direct connection, and the like. Such a communication interface may be one or more of the following: any type of network interface (eg, network interface card (NIC)), wired or wireless (such as IEEE 802.11 wireless LAN (WLAN)) wireless interface, global microwave Access Interoperability (Wi-MAX) interface, Ethernet interface, Universal Serial Bus (USB) interface, cellular network interface, Bluetooth ^™ interface, Near Field Communication (NFC) interface, and the like. Communication interface 606 may facilitate communication within a variety of networks and protocol types, including wired networks (eg, LAN, cable, etc.) and wireless networks (eg, WLAN, cellular, satellite, etc.), the Internet, and the like. Communication interface 606 may also provide for communication with external storage devices (not shown) such as in storage arrays, network attached storage, storage area networks, and the like.

In some examples, a display device 608, such as a monitor, may be included for displaying information and images to a user. Other I/O devices 610 may be devices that receive various inputs from the user and provide various outputs to the user, and may include touch input devices, gesture input devices, cameras, keyboards, remote controls, mice, printers, audio input/ output devices, etc.

The techniques described herein may be supported by these various configurations of computer device 600 and are not limited to the specific examples of the techniques described herein. For example, this functionality can also be implemented in whole or in part on the "cloud" using a distributed system. The cloud includes and/or represents a platform for resources. The platform abstracts the underlying functionality of the cloud's hardware (eg, servers) and software resources. Resources may include applications and/or data that may be used when computing processing is performed on servers remote from computer device 600 . Resources may also include services provided over the Internet and/or over subscriber networks such as cellular or Wi-Fi networks. The platform may abstract resources and functions to connect computer device 600 with other computer devices. Thus, the implementation of the functions described herein can be distributed across the cloud. For example, functionality may be implemented partly on computer device 600 and partly through a platform that abstracts the functionality of the cloud.

While the present disclosure has been illustrated and described in detail in the accompanying drawings and the foregoing description, such illustration and description are to be considered illustrative and schematic and not restrictive; the present disclosure is not limited to the disclosed Example. Variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed subject matter, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps not listed, the indefinite article "a" or "an" does not exclude a plurality, and the term "a plurality" means two or more, And the term "based on" should be interpreted as "based at least in part on". The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Some exemplary aspects of the present disclosure will be described below.

Aspect 1, a voice interaction method for a vehicle cockpit, the vehicle cockpit comprising a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones being collectively configured with a current session cache for storing a current multi-round session, and Each sound region in the plurality of sound regions is configured with a corresponding sound region buffer for storing historical semantic content from the sound region within a preset time window, and the method includes:

obtaining the current semantic content from the first sound region of the plurality of sound regions;

The current multi-round session is updated based on the current session buffer, the sound area buffer of the first sound area, and the sound area buffers of other sound areas in the plurality of sound areas except the first sound area, wherein the updated current multi-round session is the same as the associated with the current semantic content; and

Process the current semantic content associated with the updated current multi-round session.

Aspect 2, the method of aspect 1, wherein updating the current multi-round session includes:

determining whether historical semantic content is stored in the sound region cache of the first sound region;

determining whether the current semantic content and the historical semantic content belong to the same multi-round session in response to determining that historical semantic content is stored in the tone zone cache of the first tone zone; and

In response to determining that the current semantic content and the historical semantic content belong to the same multi-round session, the current semantic content and the historical semantic content are stored in the current session cache to update the current multi-round session.

Aspect 3, the method of aspect 2, further comprising:

In response to determining that the current semantic content and the historical semantic content do not belong to the same multi-round session, determining whether the current semantic content and the current multi-round session belong to the same multi-round session; and

In response to determining that the current semantic content and the current multi-round session belong to the same multi-round session, the current semantic content and the current multi-round session are stored in the current session cache to update the current multi-round session.

Aspect 4, the method of aspect 3, further comprising:

In response to determining that the current semantic content and the current multi-round session do not belong to the same multi-round session, determining the current semantic content and the historical semantic content stored in the voice zone cache of the remaining voice zones in the plurality of voice zones except the first voice zone belong to the same multi-round session; and

In response to determining that the current semantic content and the historical semantic content stored in the sound region cache of the second sound region of the plurality of sound regions belong to the same multi-round session, the current semantic content is compared with the historical semantic content stored in the sound region cache of the second sound region. The stored historical semantic content is stored together in the current session cache to update the current multi-round session, where the second pitch is a different pitch than the first.

Aspect 5, the method of aspect 4, further comprising:

In response to determining that the current semantic content does not belong to the same multi-round session as the historical semantic content stored in the sound region caches of the remaining sound regions of the plurality of sound regions except the first sound region, storing the current semantic content in the current session cache to update the current multi-round session.

Aspect 6, the method of aspect 2, further comprising:

determining whether the current semantic content and the current multi-round session belong to the same multi-round conversation in response to determining that the historical semantic content is not stored in the sound region cache of the first sound region; and

In response to determining that the current semantic content and the current multi-round session belong to the same multi-round session, the current semantic content and the current multi-round session are stored in the current session cache to update the current multi-round session.

Aspect 7, the method of aspect 6, further comprising:

In response to determining that the current semantic content and the current multi-round session do not belong to the same multi-round session, determining the current semantic content and the historical semantic content stored in the voice zone cache of the remaining voice zones in the plurality of voice zones except the first voice zone belong to the same multi-round session; and

In response to determining that the current semantic content and the historical semantic content stored in the sound region cache of the second sound region of the plurality of sound regions belong to the same multi-round session, the current semantic content is compared with the historical semantic content stored in the sound region cache of the second sound region. The stored historical semantic content is stored together in the current session cache to update the current multi-round session, where the second pitch is a different pitch than the first.

Aspect 8, the method of aspect 7, further comprising:

In response to determining that the current semantic content does not belong to the same multi-round session as the historical semantic content stored in the sound region caches of the remaining sound regions of the plurality of sound regions except the first sound region, storing the current semantic content in the current session cache to update the current multi-round session.

Aspect 9, the method of aspect 1, further comprising:

The current semantic content is stored in the sound region cache of the first sound region to update the historical semantic content stored in the sound region cache of the first sound region.

Aspect 10, a voice interaction device for a vehicle cabin, the vehicle cabin comprising a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones are collectively configured with a current session cache for storing a current multi-round session, and Each sound area in the plurality of sound areas is configured with a corresponding sound area buffer for storing historical semantic content from the sound area within a preset time window, and the device includes:

an obtaining module, configured to obtain the current semantic content of the first sound region from the plurality of sound regions;

The updating module is configured to update the current multi-round session based on the current session cache, the sound region cache of the first sound region and the sound region caches of other sound regions in the plurality of sound regions except the first sound region, wherein after the update is associated with the current semantic content; and

A processing module configured to process the current semantic content associated with the updated current multi-round session.

Aspect 11, a computer device for a vehicle cabin, the vehicle cabin comprising a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones are collectively configured with a current session cache for storing a current multi-round session, and a plurality of Each of the sound regions is configured with a corresponding sound region buffer for storing historical semantic content from the sound region within a preset time window, and the computer equipment includes:

at least one processor; and

at least one memory having a computer program stored thereon,

Wherein, the computer program, when executed by at least one processor, causes at least one processor to execute the method of any one of aspects 1-9.

Aspect 12, a vehicle comprising the voice interaction device of aspect 10 or the computer equipment of aspect 11.

Aspect 13, a computer-readable storage medium having stored thereon a computer program that, when executed by a processor, causes the processor to perform the method of any one of aspects 1-9.

Aspect 14, a computer program product comprising a computer program which, when executed by a processor, causes the processor to perform the method of any of aspects 1-9.

Claims (10)

1. A voice interaction method for a vehicle cockpit comprising a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones being jointly configured with a current session cache for storing the current multi-round conversation, and each sound region in the plurality of sound regions is configured with a corresponding sound region buffer for storing historical semantic content from the sound region within a preset time window, the method includes: obtaining the current semantic content from the first sound region of the plurality of sound regions; updating the current multi-round session based on the current session buffer, the tone buffer of the first tone area, and the tone buffers of the remaining tone areas of the plurality of tone areas except the first tone area, wherein the updated current multi-round conversation is associated with the current semantic content; and The current semantic content associated with the updated current multi-round session is processed. 2. The method of claim 1, wherein updating the current multi-round session comprises: Determine whether the historical semantic content is stored in the tone cache of the first tone area; determining whether the current semantic content and the historical semantic content belong to the same multi-round session in response to determining that historical semantic content is stored in the tone zone cache of the first tone zone; and In response to determining that the current semantic content and the historical semantic content belong to the same multi-round session, storing the current semantic content with the historical semantic content in the current session cache to update the current multi-round session . 3. The method of claim 2, further comprising: In response to determining that the current semantic content and the historical semantic content do not belong to the same multi-round session, determining whether the current semantic content and the current multi-round session belong to the same multi-round session; and In response to determining that the current semantic content and the current multi-round session belong to the same multi-round session, storing the current semantic content with the current multi-round session in the current session cache to update the current multi-round session. round session. 4. The method of claim 3, further comprising: In response to determining that the current semantic content and the current multi-round conversation do not belong to the same multi-round session, determining the relationship between the current semantic content and the remaining sound regions of the plurality of sound regions other than the first sound region Whether the historical semantic content stored in the sound zone cache belongs to the same multi-round session; and In response to determining that the current semantic content belongs to the same multi-round session as the historical semantic content stored in the voice zone cache of a second voice zone of the plurality of voice zones, comparing the current semantic content with the second voice zone The historical semantic content stored in the tone area cache of the tone area is stored together in the current session cache to update the current multi-round conversation, wherein the second tone area is different from the first tone area sound area. 5. The method of claim 4, further comprising: In response to determining that the current semantic content does not belong to the same multi-round session as the historical semantic content stored in the sound region caches of the remaining sound regions of the plurality of sound regions except the first sound region, the The current semantic content is stored in the current session cache to update the current multi-round session. 6. The method of claim 2, further comprising: determining whether the current semantic content and the current multi-round session belong to the same multi-round session in response to determining that historical semantic content is not stored in the voice zone cache of the first voice zone; and In response to determining that the current semantic content and the current multi-round session belong to the same multi-round session, storing the current semantic content with the current multi-round session in the current session cache to update the current multi-round session. round session. 7. The method of claim 6, further comprising: In response to determining that the current semantic content and the current multi-round conversation do not belong to the same multi-round session, determining the relationship between the current semantic content and the remaining sound regions of the plurality of sound regions other than the first sound region Whether the historical semantic content stored in the sound zone cache belongs to the same multi-round session; and In response to determining that the current semantic content belongs to the same multi-round session as the historical semantic content stored in the voice zone cache of a second voice zone of the plurality of voice zones, comparing the current semantic content with the second voice zone The historical semantic content stored in the tone area cache of the tone area is stored together in the current session cache to update the current multi-round conversation, wherein the second tone area is different from the first tone area sound area. 8. The method of claim 7, further comprising: In response to determining that the current semantic content does not belong to the same multi-round session as the historical semantic content stored in the sound region caches of the remaining sound regions of the plurality of sound regions except the first sound region, the The current semantic content is stored in the current session cache to update the current multi-round session. 9. The method of claim 1, further comprising: The current semantic content is stored in the sound region cache of the first sound region to update the historical semantic content stored in the sound region cache of the first sound region. 10. A voice interaction device for a vehicle cockpit, the vehicle cockpit comprising a plurality of sound zones respectively corresponding to a plurality of vehicle seats, the plurality of sound zones being jointly configured with a current session cache for storing a current multi-round conversation, and each sound region in the plurality of sound regions is configured with a corresponding sound region buffer for storing historical semantic content from the sound region within a preset time window, the device includes: an acquisition module, the acquisition module is configured to acquire the current semantic content from the first sound region in the plurality of sound regions; an update module, the update module is configured to be based on the current session cache, the tone cache of the first tone zone, and the tone zones of the remaining tone zones in the plurality of tone zones except the first tone zone caching, updating the current multi-round session, wherein the updated current multi-round session is associated with the current semantic content; and A processing module configured to process the current semantic content associated with the updated current multi-round session.

Images