Vehicle control method and device, control equipment, vehicle and storage medium

Abstract

The embodiment of the invention provides a vehicle control method and device, control equipment, a vehicle and a storage medium. The method comprises the following steps: acquiring global voice data in a cabin; performing voice recognition on the global voice data, and determining an actually measured relativity between a called word and a speaker and a called person in the global voice data; performing identity recognition based on the name word and the actually measured relativity, and determining a speaker target identity corresponding to the speaker and a called person target identity corresponding to the called person; a target component of the vehicle is controlled to perform a personalized service based on the speaker target identity and the called person target identity. According to the method, non-inductive accurate identity recognition can be carried out on the user through the name word and the relative relationship in the vehicle driving process of the user, then the vehicle is controlled to carry out accurate personalized service on the user, the user does not need to actively carry out identity recognition, the user identity can be accurately recognized, and the user experience is improved when the personalized service is provided for the user. And the driving experience and the driving safety performance of the user are improved.

Description

Technical Field

[0001] This application relates to the field of smart cockpit user identification technology, and in particular to a vehicle control method, device, control equipment, vehicle, and storage medium. Background Technology

[0002] With the continuous development of vehicle intelligence technology, users typically control their vehicles personalized via voice commands. Current technology requires user authentication through inputting verification codes or facing a camera before personalized control can be implemented. This method can distract drivers, impacting safety, and the cumbersome operation negatively affects the user's driving experience. Therefore, improving the user's driving experience and safety during vehicle control is a pressing technical challenge. Summary of the Invention

[0003] This application provides a vehicle control method, device, control equipment, vehicle, and storage medium, aiming to solve the technical problem of how to improve the user's driving experience and driving safety performance during vehicle control.

[0004] A vehicle control method, comprising: Acquire full-area voice data within the cockpit; Speech recognition is performed on the full-domain speech data to determine the address words in the full-domain speech data and the measured kinship relationship between the speaker and the person being addressed; Based on the address terms and the measured kinship relationships, identity recognition is performed to determine the speaker's target identity and the person being addressed's target identity. Based on the speaker's target identity and the called person's target identity, the target components of the vehicle are controlled to perform personalized services.

[0005] In this embodiment, full-domain voice data is acquired and analyzed in real time. During the user's driving experience, the system can determine the terms of address and the measured kinship relationship between the speaker and the person being addressed in the full-domain voice data without the user's awareness. Based on these terms of address and measured kinship relationships, identity recognition is performed, accurately determining the speaker's target identity and the person being addressed's target identity. This eliminates the need for the user to actively perform complex identity recognition actions; identity recognition is directly based on the terms of address generated during the user's daily interactions within the vehicle and the measured kinship relationships. This achieves natural and seamless identity recognition for the user. During the user's driving experience, the system naturally and seamlessly determines the personalized services corresponding to the speaker's target identity and the person being addressed, controlling the vehicle's target components to provide precise personalized services to the speaker and the person being addressed. This method achieves natural and accurate identity recognition without the user's awareness by using the address and kinship relationship in the full-domain voice data of the user in the cockpit. Based on the target identities of the speaker and the person being addressed, it accurately provides personalized services to the user without requiring the user to perform complex identity authentication operations. This method does not distract the user's attention during the driving process, improving the user's driving experience and the safety performance of the vehicle.

[0006] Preferably, the step of identifying the speaker's target identity based on the address term and the measured kinship relationship, and determining the addressee's target identity, includes: Based on the address, the identity of the person being addressed is matched to determine the target identity of the person being addressed. Based on the target identity of the person being addressed and the measured kinship relationship, identity recognition is performed to determine the speaker's target identity.

[0007] In this embodiment, the identity of the person being addressed is matched by the term of address to determine the target identity of the person being addressed. Based on the target identity of the person being addressed and the measured kinship relationship, the speaker's target identity is determined. This method does not require any operation from the user in the cabin and can determine the speaker's target identity and the person being addressed's target identity in a relatively natural way, thereby improving the user's driving experience and driving safety performance.

[0008] Preferably, the step of matching the identity of the person being addressed based on the address term to determine the target identity of the person being addressed includes: The query is based on the address term and a pre-stored first mapping table, which stores the mapping relationship between address terms and user identities. When no addressing term exists in the first mapping table, the measured basic features of the person being addressed are obtained, and the pre-stored second mapping table is queried based on the measured basic features. The user identity that matches the measured basic features in the second mapping table is determined as the target identity of the person being addressed. The second mapping table is used to store the mapping relationship between user identity and its corresponding basic features and common locations. When a title exists in the first mapping table, the user identity corresponding to the title in the first mapping table is determined as the initial identity of the person being addressed. Based on the initial identity of the person being addressed, the second mapping table is queried to determine the first basic feature corresponding to the initial identity of the person being addressed. Based on the measured basic feature of the person being addressed and the first basic feature, the initial identity of the person being addressed is verified to determine the target identity of the person being addressed.

[0009] In this embodiment, different methods are used to accurately determine the target identity of the person being addressed, depending on whether a title exists in the first mapping table. When a title exists in the first mapping table, the user identity in the second mapping table that matches the measured basic features is determined as the target identity of the person being addressed. When a title exists in the first mapping table, the initial identity of the person being addressed is determined, and the initial identity of the person being addressed is verified based on the measured basic features corresponding to the person being addressed and the first basic features corresponding to the initial identity of the person being addressed. This method can accurately determine the target identity of the person being addressed, and it does not require the person being addressed to perform any identity verification operation. It can naturally and seamlessly determine the target identity of the person being addressed, improving the driving experience and driving safety performance of the person being addressed.

[0010] Preferably, the measured basic characteristic includes measured weight, and the first basic characteristic includes a first weight range; The step of querying the second mapping table based on the initial identity of the person being addressed to determine the first basic feature corresponding to the initial identity of the person being addressed, and performing identity verification on the initial identity of the person being addressed based on the measured basic feature of the person being addressed and the first basic feature to determine the target identity of the person being addressed, includes: Based on the initial identity of the person being addressed, the second mapping table is queried to determine the first weight range and the first frequently used location corresponding to the initial identity of the person being addressed. If the measured weight of the person being addressed at the first commonly used location is within the first weight range, then the initial identity of the person being addressed is determined as the target identity of the person being addressed. If the measured weight of the person being addressed at the first commonly used location is not within the first weight range, and the measured weight of the person being addressed at the remaining locations is within the first weight range, obtain the voice response of the person being addressed. Based on the voice response of the person being addressed, verify the initial identity of the person being addressed and determine the target identity of the person being addressed.

[0011] In this embodiment, by comparing the measured weight at the first commonly used location with a first weight range, the target identity of the person being addressed can be quickly determined. If the measured weight at the first commonly used location is not within the first weight range, the measured weight at the remaining locations is further judged to see if it is within the first weight range. Combined with the voice response of the person being addressed, the remaining locations where the measured weight is within the first weight range are verified to accurately determine the target identity of the person being addressed.

[0012] Preferably, the first basic feature further includes a first facial feature; The process of verifying the initial identity of the person being addressed based on their voice response, and determining the target identity of the person being addressed, includes: Based on the voice response of the person being addressed, the location of the person being addressed is determined, and the actual measured location corresponding to the person being addressed is determined. If the measured location is the same as the remaining location, then the initial identity of the person being called is determined as the target identity of the person being called; If the measured location is not the same as the remaining location, the measured facial features corresponding to the measured location are obtained. If the measured facial features are consistent with the first facial features, the initial identity of the person being called is determined as the target identity of the person being called.

[0013] In this embodiment, the measured location is first matched with the remaining locations corresponding to the measured weight within the first weight range. If the match is successful, the initial identity of the person being called is directly determined as the target identity of the person being called, which is relatively efficient and convenient. If the match fails, the measured facial features are matched with the first facial features to accurately determine the target identity of the person being called. This method can effectively save computing power and achieve natural and seamless identity verification of the person being called.

[0014] Preferably, the step of identifying the speaker's target identity based on the target identity of the person being addressed and the measured kinship relationship includes: Based on the target identity of the person being addressed and the measured kinship, the pre-stored kinship network is queried to determine the initial speaker identity corresponding to the speaker; Based on the speaker's initial identity, a pre-stored second mapping table is queried to determine the second basic feature and second frequently used position corresponding to the speaker's initial identity. The second mapping table is used to store the mapping relationship between the user identity and its corresponding basic feature and frequently used position. Based on the second basic feature and second common location corresponding to the speaker's initial identity, the speaker's initial identity is verified to determine the speaker's target identity.

[0015] In this embodiment, the initial speaker identity is determined based on the target identity of the person being addressed and the measured kinship relationship. Then, according to a pre-stored second mapping table, the initial speaker identity is verified using the second basic feature and second frequently used location corresponding to the initial speaker identity, thus accurately determining the speaker's target identity. This method eliminates the need for user intervention within the cabin, naturally determining the speaker's target identity and improving the user's driving experience and safety.

[0016] A vehicle control device, comprising: Acquire full-area voice data within the cockpit; Speech recognition is performed on the full-domain speech data to determine the address words in the full-domain speech data and the measured kinship relationship between the speaker and the person being addressed; Based on the address terms and the measured kinship relationships, identity recognition is performed to determine the speaker's target identity and the person being addressed's target identity. Based on the speaker's target identity and the called person's target identity, the target components of the vehicle are controlled to perform personalized services.

[0017] A control device includes a processor and a memory, wherein, Memory, used to store computer programs; The processor is used to execute the program stored in the memory to implement the vehicle control method described above.

[0018] A vehicle including the aforementioned control equipment.

[0019] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the vehicle control method described above. Attached Figure Description

[0020] Figure 1 This is a flowchart of a vehicle control method provided in an embodiment of this application; Figure 2 This is another flowchart of a vehicle control method provided in one embodiment of this application; Figure 3 This is another flowchart of a vehicle control method provided in one embodiment of this application; Figure 4 This is another flowchart of a vehicle control method provided in one embodiment of this application; Figure 5 This is another flowchart of a vehicle control method provided in one embodiment of this application; Figure 6 This is another flowchart of a vehicle control method provided in one embodiment of this application; Figure 7 This is a schematic diagram of a vehicle control device provided in an embodiment of this application; Figure 8 This is a structural diagram of the control device provided in the embodiments of this application. Detailed Implementation

[0021] To make the technical problems, technical solutions, and beneficial effects solved by this application clearer, the following detailed description is provided in conjunction with embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0022] This application provides a vehicle control method, including: acquiring full-domain voice data within the cockpit; This method performs speech recognition on full-domain speech data to determine the address words and the measured kinship relationship between the speaker and the person being addressed. Based on the address words and the measured kinship relationship, it performs identity recognition to determine the speaker's target identity and the person being addressed's target identity. Based on the speaker's target identity and the person being addressed's target identity, it controls the target components of the vehicle to execute personalized services. This method can achieve seamless and accurate user identity recognition during the user's driving process through address words and kinship relationships, thereby controlling the vehicle to provide precise personalized services. This method can accurately identify the user's identity without the user's active identification, improving the user's driving experience and driving safety while providing precise personalized services.

[0023] In one embodiment, such as Figure 1 As shown, a vehicle control method is provided, which is applied to... Figure 8 Taking the control equipment in the example, the following steps are included: S101: Acquire full-domain voice data within the cockpit; S102: Perform speech recognition on the full-domain speech data to determine the address words in the full-domain speech data and the measured kinship relationship between the speaker and the person being addressed; S103: Based on the address words and the measured kinship, perform identity recognition to determine the speaker target identity corresponding to the speaker and the addressee target identity corresponding to the addressee; S104: Based on the speaker's target identity and the called person's target identity, control the target components of the vehicle to perform personalized services.

[0024] Among them, full-domain voice data refers to the voice data of users in the cockpit.

[0025] As an example, in step S101, the control device receives real-time voice data transmitted by the omnidirectional microphone array inside the vehicle cabin as omnidirectional voice data, which facilitates more natural identity authentication of users inside the cabin. In this example, the vehicle cabin is equipped with an omnidirectional microphone array covering the driver's seat, front passenger seat, left rear seat, and right rear seat. The omnidirectional microphone array collects voice data from users inside the vehicle cabin in real time and transmits the voice data to the vehicle's control device. The control device identifies the received voice data as omnidirectional voice data. This process does not require active user participation, allowing for natural and seamless user identity authentication based on the omnidirectional voice data.

[0026] In this context, "address term" refers to the words corresponding to natural forms of address, such as "Li XX," "Dad," and "Son." "Speaker" refers to the person uttering the address term, and "received person" refers to the person indicated by the address term. "Measured kinship relationship" refers to the kinship relationship between the recipient and the speaker, such as father-son, mother-son, husband-wife, and grandparent-grandchild relationships.

[0027] As an example, in step S102, the control device performs speech recognition on the real-time received global speech data. Specifically, this includes recognizing address words in the global speech data, identifying address words in the global speech data, and determining the measured kinship relationship between the person being addressed and the speaker who issued the address word based on the address word. For example, if the global speech data is "Dad, I'm a little cold," the control device recognizes the address word as "Dad." Based on the address word "Dad," the control device analyzes the kinship relationship between the person being addressed and the speaker who issued the address word, determining that the measured kinship relationship between the person being addressed and the speaker is father and son. In this example, the measured kinship relationship is further determined only after it is determined that the global speech data contains an address word. If it is determined that the global speech data does not contain an address word, the global speech data is discarded, and the subsequent measured kinship relationship determination is not performed, avoiding invalid calculations and saving system computing power.

[0028] Here, the speaker's target identity refers to the user identity corresponding to the speaker obtained through identity recognition. The addressed person's target identity refers to the user identity corresponding to the addressed person obtained through identity recognition.

[0029] As an example, in step S103, the control device identifies the speaker and the person being addressed based on the address and the measured kinship relationship, thus identifying the speaker's target identity and the person being addressed's target identity. This method allows for natural and seamless user identity recognition during vehicle use, without requiring the speaker and the person being addressed to actively undergo complex user authentication. It accurately identifies the speaker's target identity and the person being addressed's target identity based on the address and the measured kinship relationship, improving the user's driving experience and safety.

[0030] Personalized services refer to the services that users require. For example, while driving, users may need to control the vehicle's temperature control system to provide their preferred temperature and control the vehicle's audio-visual entertainment system to play their preferred music, among other personalized services.

[0031] As an example, in step S104, after determining the speaker's target identity and the called person's target identity, the control device queries the system database for the speaker's target identity and the identity service mapping table pre-stored by the personalized service memory module to determine the personalized services corresponding to the speaker's target identity and the called person's target identity. The control device then controls the target components of the vehicle to execute these personalized services. In this example, the identity service mapping table stores the mapping relationship between user identities and personalized services. For instance, if the control device finds that the personalized service corresponding to the speaker's target identity is a preferred temperature of 26°C and the personalized service corresponding to the called person's target identity is a preference for classical music, it determines that the target components are the temperature control component of the temperature control system and the audio component of the audio-visual entertainment system, respectively. The control device then controls the temperature control component of the temperature control system to adjust the cabin temperature to 26°C and controls the audio component of the audio-visual entertainment system to play classical music. This method determines and executes the personalized services corresponding to the speaker's target identity and the called person's target identity in a natural and seamless manner, enhancing the user's personalized driving experience and driving safety.

[0032] In this embodiment, full-domain voice data is acquired and analyzed in real time. During the user's driving experience, the system can determine the terms of address and the measured kinship relationship between the speaker and the person being addressed in the full-domain voice data without the user's awareness. Based on these terms of address and measured kinship relationships, identity recognition is performed, accurately determining the speaker's target identity and the person being addressed's target identity. This eliminates the need for the user to actively perform complex identity recognition actions; identity recognition is directly based on the terms of address generated during the user's daily interactions within the vehicle and the measured kinship relationships. This achieves natural and seamless identity recognition for the user. During the user's driving experience, the system naturally and seamlessly determines the personalized services corresponding to the speaker's target identity and the person being addressed, controlling the vehicle's target components to provide precise personalized services to the speaker and the person being addressed. This method achieves natural and accurate identity recognition without the user's awareness by using the address and kinship relationship in the full-domain voice data of the user in the cockpit. Based on the target identities of the speaker and the person being addressed, it accurately provides personalized services to the user without requiring the user to perform complex identity authentication operations. This method does not distract the user's attention during the driving process, improving the user's driving experience and the safety performance of the vehicle.

[0033] In one embodiment, such as Figure 2 As shown, step S103, which involves identity recognition based on the address and the measured kinship relationship to determine the speaker's target identity and the person being addressed's target identity, includes: S201: Match the identity of the person being addressed based on the address words to determine the target identity of the person being addressed; S202: Based on the target identity of the person being addressed and the measured kinship relationship, identify the speaker's target identity.

[0034] As an example, in step S201, the control device queries a pre-stored mapping table that stores address terms and user identities to determine the initial identity of the person being addressed corresponding to the address term. Then, in the pre-stored mapping table that stores user identities and basic features, it queries the basic features corresponding to the initial identity of the person being addressed and obtains the measured basic features corresponding to the person being addressed. If the measured basic features match the corresponding basic features in the mapping table, the initial identity of the person being addressed is determined as the target identity. If the measured basic features do not match the corresponding basic features in the mapping table, the device queries the pre-stored mapping table that stores user identities and basic features to see if there is a basic feature that matches the measured basic features corresponding to the person being addressed. If it exists, the user identity corresponding to that basic feature in the mapping table is determined as the target feature of the person being addressed. If it does not exist, the user identity of the person being addressed is determined as a new user, a new user identity is set for the person being addressed, and the new user identity and the measured basic features corresponding to the person being addressed are stored in the mapping table that stores user identities and basic features, thereby enriching the mapping table for storing user identities and basic features and improving the efficiency of subsequent identity recognition. Here, "initial identity of the person being addressed" refers to the user identity initially identified. "Basic features" refers to the characteristics corresponding to a user's identity, used to distinguish users with different identities. "User identity" refers to a user's identity, such as car owner A, user B, and user C, used to characterize different users. "Measured basic features" refers to the basic features actually detected.

[0035] As an example, in step S202, the control device queries the kinship relationships between different user identities stored in the pre-stored database, determines the user identity corresponding to the target identity of the person being addressed and the measured kinship relationship, and determines this user identity as the speaker's initial identity. In a pre-stored mapping table used to store user identities and basic features, the device queries the basic features corresponding to the speaker's initial identity and obtains the measured basic features corresponding to the speaker's initial identity in real time. If the measured basic features match the corresponding basic features in the mapping table, the speaker's initial identity is determined as the speaker's target identity. If the measured basic features do not match the corresponding basic features in the mapping table, the device queries the pre-stored mapping table used to store user identities and basic features to see if there is a basic feature that matches the speaker's measured basic features. If it exists, the user identity corresponding to that basic feature in the mapping table is determined as the speaker's target identity. If it does not exist, the speaker's user identity is determined as a new user, a new user identity is set for the speaker, and the new user identity and the corresponding measured basic features of the speaker are stored in the mapping table used to store user identities and basic features, thereby enriching the mapping table used to store user identities and basic features and improving the efficiency of subsequent identity recognition. The initial speaker identity refers to the user identity of the speaker as initially identified.

[0036] In this embodiment, the identity of the person being addressed is matched by the term of address to determine the target identity of the person being addressed. Based on the target identity of the person being addressed and the measured kinship relationship, the speaker's target identity is determined. This method does not require any operation from the user in the cabin and can determine the speaker's target identity and the person being addressed's target identity in a relatively natural way, thereby improving the user's driving experience and driving safety performance.

[0037] In one embodiment, such as Figure 3 As shown, step S201, which involves matching the identity of the person being addressed based on the address term to determine the target identity of the person being addressed, includes: S301: Query the pre-stored first mapping table based on the title word. The first mapping table is used to store the mapping relationship between the title word and the user identity. S302: When there is no addressing term in the first mapping table, obtain the measured basic features of the person being addressed, query the pre-stored second mapping table based on the measured basic features, and determine the user identity in the second mapping table that matches the measured basic features as the target identity of the person being addressed; the second mapping table is used to store the mapping relationship between user identity and its corresponding basic features and common locations. S303: When a title exists in the first mapping table, the user identity corresponding to the title in the first mapping table is determined as the initial identity of the person being addressed. Based on the initial identity of the person being addressed, the second mapping table is queried to determine the first basic feature corresponding to the initial identity of the person being addressed. Based on the measured basic feature and the first basic feature of the person being addressed, the initial identity of the person being addressed is verified to determine the target identity of the person being addressed.

[0038] The first mapping table stores the mapping relationship between at least one title and a user identity. Understandably, due to the diversity of family roles a user identity assumes, a user identity may correspond to more than one title. Therefore, a title set contains at least one title, and one user identity corresponds to one title set, in order to accurately determine the user identity based on the title set. For example, as shown in Table 1 below, the titles corresponding to the user identity "Car Owner A" include "Dad," "Husband," and "Zhang X," etc.

[0039] As an example, in step S301, after obtaining the address terms from the global speech data, the control device queries the address term set corresponding to each user identity in the pre-stored first mapping table to determine whether there exists an address term set containing the address terms from the global speech data, and then determines whether the address terms from the global speech data exist in the first mapping table. In this example, the control device queries the first mapping table in Table 1 based on the address terms from the global speech data to determine whether there is an address term set containing the address terms from the global speech data. If it is determined that there is an address term set containing the address terms from the global speech data, it is determined that the address terms from the global speech data exist in the first mapping table. If it is determined that there is no address term set containing the address terms from the global speech data, it is determined that the address terms from the global speech data do not exist in the first mapping table. In this example, the first mapping table is stored in the memory module of the system database.

[0040] Table 1 The second mapping table stores the mapping relationship between user identities, their corresponding basic characteristics, and frequently used locations. Frequently used locations refer to the positions a user often sits in the vehicle, which can be obtained from historical data. As shown in Table 2, when user identity is vehicle owner A, frequently used locations include the driver's seat and the front passenger seat; when user identity is user B, frequently used locations include the left rear seat; when user identity is user C, frequently used locations include the front passenger seat and the right rear seat; and when user identity is user D, frequently used locations include the right rear seat.

[0041] As an example, in step S302, when the control device determines that there is no recognized address term in the first mapping table, it acquires the voice response of the person being addressed, locates the person's position based on the voice response, determines the location of the person being addressed, and acquires the measured basic features corresponding to the location of the person being addressed. Table 2 below shows the second mapping table in this example, which stores user identities and basic features. In this example, the control device matches the measured basic features with the basic features corresponding to each user identity in the pre-stored second mapping table. If a basic feature in the mapping table matches the measured basic feature, the user identity corresponding to that basic feature is determined as the target identity of the person being addressed. If no basic feature in the mapping table matches the measured basic feature, it indicates that there is no user identity corresponding to the person being addressed in the second mapping table. When the control device determines that there is no user identity corresponding to the person being addressed, it sets a new user identity for that person and adds the new user identity, measured basic features, and the location of the person being addressed to the second mapping table to complete the second mapping table.

[0042] In this example, as shown in Table 2, the basic features include weight range and facial features. The measured basic features include measured weight and measured facial features. Measured weight refers to the actual weight measured. Measured facial features refer to the facial features obtained through actual testing. In this example, the vehicle cabin is equipped with an identity verification module, which includes a seat pressure sensor and a camera device installed at each location to assist in identity verification. The seat pressure sensor at each location acquires the measured weight at that location and transmits it to the control device. The camera device at each location acquires the measured facial features of the user at that location and transmits the measured facial features to the control device.

[0043] In this example, the control device acquires measured basic features, including measured weight and measured facial features. It first determines whether a weight range containing the measured weight exists in the second mapping table. For example, if the measured weight is 15KG, querying the second mapping table determines that when the user's identity is User B, the weight range is 15±2KG, which includes the measured weight. As another example, if the measured weight is 80KG, querying the second mapping table determines that the weight range corresponding to all user identities does not include the measured weight of the person being addressed.

[0044] When the control device determines that there is no weight range containing the measured weight in the second mapping table, it determines that the second mapping table does not contain the user identity corresponding to the person being addressed. Since the user identity corresponding to the person being addressed is a new identity, the user identity of the person being addressed is set to the new user, and the address, measured weight, and measured facial features corresponding to the person being addressed are added to the mapping relationship of the new user in the second mapping table to improve the mapping table and facilitate more comprehensive and accurate identification of different user identities in the future.

[0045] When the control device determines that there is a weight range containing the measured weight in the second mapping table, it determines the user identity corresponding to the weight range as the initial identity of the person being addressed. It then matches the facial features corresponding to the initial identity of the person being addressed with the measured facial features and determines whether the facial features corresponding to the initial identity of the person being addressed are consistent with the measured facial features. If they are consistent, the facial matching result is determined to be a successful match; otherwise, the facial matching result is determined to be a failed match.

[0046] When the control device determines that the facial matching result is a successful match, it adds the address words recognized in the full-domain speech data to the address word set corresponding to the initial identity of the person being addressed in the first mapping table, and determines the initial identity of the person being addressed as the target identity of the person being addressed. Understandably, if there exists a weight range that matches the measured weight, and the facial matching result is a successful match, it indicates that the weight range in the initial identity of the person being addressed matches the measured weight of the person being addressed, and the facial features in the initial identity of the person being addressed also match the measured facial features of the person being addressed. That is, the basic features of the person being addressed match the basic features of the initial identity of the person being addressed. Therefore, the initial identity of the person being addressed is the target identity of the person being addressed.

[0047] When the control device determines that the facial matching result is a failure, it establishes a new user identity in the second mapping table. This involves adding the address words, measured weight, and measured facial features identified in the full-domain voice data to the new user's mapping relationship in the second mapping table. Understandably, if the facial matching result is a failure, it indicates that the user identity of the person being addressed is not present in the second mapping table. Therefore, in the second mapping table, a new user identity is set for the person being addressed, and the address words, measured weight, and measured facial features identified in the full-domain voice data are added to the new user's mapping relationship in the second mapping table. This completes the mapping table, enriching the user identity and corresponding basic feature data, thus ensuring the convenience of user identification and authentication during subsequent vehicle use.

[0048] Table 2 Among them, the first basic feature refers to the basic feature in the second mapping table that corresponds to the user's initial identity as the person being addressed.

[0049] As an example, in step S303, when the control device determines that a set of address words containing the identified address words exists in the second mapping table, it determines that the address words exist in the first mapping table. The user identity corresponding to the address word set to which the address word belongs is determined as the initial identity of the person being addressed. The basic feature corresponding to the initial identity of the person being addressed in the second mapping table is determined as the first basic feature. The voice response of the person being addressed is obtained, and the location of the person being addressed is determined based on the voice response. The measured basic feature corresponding to the location of the person being addressed is obtained. The measured basic feature is matched with the first basic feature to determine whether the measured basic feature and the first basic feature are consistent. If the measured basic feature and the first basic feature are consistent, the initial identity of the person being addressed is determined as the target identity of the person being addressed. If the measured basic feature and the first basic feature are inconsistent, the second mapping table is queried to determine whether there is a basic feature in the second mapping table that is consistent with the measured basic feature. The user identity corresponding to the basic feature that is consistent with the measured basic feature is determined as the target identity of the person being addressed. If there is no basic feature in the second mapping table that matches the measured basic feature, it means that there is no user identity corresponding to the person being called in the second mapping table. A new user identity is set for the person being called, and the new user identity, the measured basic feature, and the location of the person being called are added to the second mapping table to improve the second mapping table.

[0050] In this example, the first mapping table and the second mapping table can be two mapping tables as shown in Table 1 and Table 2, or they can be a single table that combines Table 1 and Table 2 and includes a set of titles, user identity, basic characteristics, and common locations. The specific table can be determined based on the storage conditions of the system database.

[0051] In this embodiment, different methods are used to accurately determine the target identity of the person being addressed, depending on whether a title exists in the first mapping table. When a title exists in the first mapping table, the user identity in the second mapping table that matches the measured basic features is determined as the target identity of the person being addressed. When a title exists in the first mapping table, the initial identity of the person being addressed is determined, and the initial identity of the person being addressed is verified based on the measured basic features corresponding to the person being addressed and the first basic features corresponding to the initial identity of the person being addressed. This method can accurately determine the target identity of the person being addressed, and it does not require the person being addressed to perform any identity verification operation. It can naturally and seamlessly determine the target identity of the person being addressed, improving the driving experience and driving safety performance of the person being addressed.

[0052] In one embodiment, the measured basic feature includes measured weight, and the first basic feature includes a first weight range.

[0053] The first weight range refers to the weight range included in the basic features corresponding to the initial identity of the person being addressed in the second mapping table. As shown in Table 2 above, each user identity in the second mapping table includes a weight range corresponding to that user identity in its basic features. For example, when the initial identity of the person being addressed is car owner A, the first weight range is 70±2kg.

[0054] In one embodiment, such as Figure 4 As shown, step S303, which involves querying the second mapping table based on the initial identity of the person being addressed to determine the first basic feature corresponding to the initial identity of the person being addressed, and verifying the initial identity of the person being addressed based on the measured basic features and the first basic feature, to determine the target identity of the person being addressed, includes: S401: Based on the initial identity of the person being addressed, query the second mapping table to determine the first weight range and the first frequently used location corresponding to the initial identity of the person being addressed; S402: If the measured weight of the person being addressed corresponding to the first commonly used position is within the first weight range, then the initial identity of the person being addressed is determined as the target identity of the person being addressed; S403: If the measured weight of the person being addressed at the first frequently used position is not within the first weight range, and the measured weight of the person being addressed at the remaining positions is within the first weight range, obtain the voice response of the person being addressed, verify the initial identity of the person being addressed based on the voice response of the person being addressed, and determine the target identity of the person being addressed.

[0055] In this context, the first frequently used position refers to the frequently used position corresponding to the initial identity of the person being addressed in the second mapping table. In this example, the basic characteristics include weight range.

[0056] As an example, in step S401, the control device queries the second mapping table to determine the weight range and common location corresponding to the user's initial identity in the second mapping table. The weight range corresponding to the user's initial identity is determined as the first weight range, and the common location corresponding to the user's initial identity is determined as the first common location, so that it is feasible to verify the user's initial identity based on the first weight range and the first common location.

[0057] As an example, in step S402, the control device obtains the measured weight corresponding to the first commonly used position of the person being called, based on their initial identity. If the measured weight corresponding to the first commonly used position is within a first weight range, and this measured weight matches the weight of a user with the initial identity of the person being called, then the user with the initial identity of the person being called is the person being called. Therefore, the initial identity of the person being called is determined as the target identity of the person being called. For example, if the initial identity of the person being called is user B, and the first commonly used position corresponding to this initial identity is the left rear row, and the first weight range is 15±2KG, the control device obtains the measured weight aKG corresponding to the left rear row. If the measured weight aKG is within the first weight range of 15±2KG, then user B is determined as the target identity of the person being called.

[0058] As an example, in step S403, when the control device determines that the measured weight is not within the first weight range, it acquires the measured weight corresponding to each remaining position in the cabin, determines whether the measured weight corresponding to each remaining position is within the first weight range, and if it determines that there are remaining positions where the measured weight is within the first weight range, it further acquires the voice response of the person being called, determines the location of the person being called, and determines whether the location of the person being called is consistent with the remaining positions where the measured weight is within the first weight range, based on whether the initial identity of the person being called can be identified as the target identity of the person being called. If it is determined that the location of the person being called is consistent with the remaining position, it is determined that the user corresponding to the initial identity of the person being called is at that remaining position, and the initial identity of the person being called is identified as the target identity of the person being called. Understandably, the remaining positions are not the commonly used locations corresponding to the initial identity of the person being called. If there are remaining positions where the measured weight meets the weight range corresponding to the initial identity of the person being called, it is necessary to further locate the person being called and determine whether the location of the person being called is at that remaining position, so as to perform double verification of the initial identity of the person being called and accurately determine the target identity of the person being called.

[0059] In this embodiment, by comparing the measured weight at the first commonly used location with a first weight range, the target identity of the person being addressed can be quickly determined. If the measured weight at the first commonly used location is not within the first weight range, the measured weight at the remaining locations is further judged to see if it is within the first weight range. Combined with the voice response of the person being addressed, the remaining locations where the measured weight is within the first weight range are verified to accurately determine the target identity of the person being addressed.

[0060] In one embodiment, the first basic feature further includes a first facial feature.

[0061] In this context, the first facial feature refers to the facial feature among the basic features corresponding to the target identity of the person being addressed in the second mapping table. As shown in Table 2 above, different user identities correspond to different facial features. Facial features are used to characterize the uniqueness of a user's identity.

[0062] In one embodiment, such as Figure 5 As shown, step S403, which involves verifying the initial identity of the person being addressed based on their voice response, and determining the target identity of the person being addressed, includes: S501: Based on the voice response of the person being addressed, locate the person being addressed and determine the actual measured location corresponding to the person being addressed; S502: If the measured location is the same as the remaining location, then the initial identity of the person being called is determined as the target identity of the person being called; S503: If the measured position is not the same as the remaining position, the measured facial features corresponding to the measured position are obtained. If the measured facial features are consistent with the first facial features, the initial identity of the person being called is determined as the target identity of the person being called.

[0063] The measured location refers to the actual position of the person being addressed within the cockpit, determined by their voice response.

[0064] As an example, in step S501, when the control device determines the remaining position of the person being addressed within the first weight range based on their measured weight, it acquires the person's voice response and locates their position based on the voice response, thus determining the person's corresponding measured position. For example, if the speaker's global voice data is "Dad, let's go to school," and the person's voice response is "Okay," the control device determines the person's corresponding measured position based on the microphone position of the "Okay" voice response. For example, if the microphone position of the "Okay" voice response is in the right rear row, then the measured position is determined to be in the right rear row.

[0065] As an example, in step S502, the control device verifies the location of the measured position and the measured weight of the person being addressed within the remaining positions within the first weight range. It determines whether the measured position and the remaining position are the same location. If they are the same location, the target identity of the person being addressed at the remaining position is determined to be the initial identity of the person being addressed. Understandably, the measured weight corresponding to the remaining position matches the first weight range corresponding to the initial identity of the person being addressed, and the person being addressed is also identified at the remaining position. Therefore, the remaining position is the location of the person being addressed, and the target identity of the person being addressed is the initial identity of the person being addressed. This method directly performs double verification of the initial identity of the person being addressed by checking whether the measured weight corresponding to the remaining position is within the first weight range corresponding to the initial identity of the person being addressed, and by checking whether the location of the person being addressed is the remaining position. This avoids misidentification when the location corresponding to the measured weight within the first weight range is not the person being addressed, and accurately determines the target identity of the person being addressed.

[0066] Among them, measured facial features refer to the facial features of the person being addressed that were actually collected.

[0067] As an example, in step S503, the control device performs position verification on the measured position and the remaining position corresponding to the measured weight within the first weight range, and determines whether the measured position and the remaining position are the same position. If they are not the same position, it indicates that although the remaining position meets the requirements of the person being called in terms of weight characteristics, the measured position and the remaining position are not the same position. Further, the measured facial features corresponding to the measured position are obtained, so as to accurately determine whether the user identity of the person being called at the measured position meets the requirements of the first facial features corresponding to the initial identity of the person being called based on the measured facial features.

[0068] The control device matches the measured facial features with the first facial features to determine if they match. If they match, it indicates that the user identity of the person being addressed at the measured location matches the initial identity of the person being addressed, and the initial identity of the person being addressed is determined as the target identity. This method serves as a supplementary verification after identity verification via weight fails, and is used to accurately determine the target identity of the person being addressed.

[0069] In this example, when the control device determines that the measured facial features are inconsistent with the first facial features, it determines that the user identity of the person being addressed at the measured location is not the initial identity of the person being addressed. When the control device determines that there is more than one initial identity of the person being addressed, it repeats steps S401 to S403 and steps S501 to S503 to determine the target identity of the person being addressed from among the multiple initial identities of the person being addressed. If the user identity of the person being addressed at the measured location does not match all the initial identities of the people being addressed, it determines that the second mapping table does not contain the user identity corresponding to the person being addressed. The user identity of the person being addressed is a new user. Therefore, the user identity corresponding to the person being addressed is set as a new user, and the address, the measured weight that matches the first weight range, the measured location, and the measured facial features are added to the mapping relationship corresponding to the new user in the second mapping table to improve the second mapping table.

[0070] In this embodiment, the measured location is first matched with the remaining locations corresponding to the measured weight within the first weight range. If the match is successful, the initial identity of the person being called is directly determined as the target identity of the person being called, which is relatively efficient and convenient. If the match fails, the measured facial features are matched with the first facial features to accurately determine the target identity of the person being called. This method can effectively save computing power and achieve natural and seamless identity verification of the person being called.

[0071] In one embodiment, such as Figure 6 As shown, step 202, which involves identity recognition based on the target identity of the person being addressed and the measured kinship relationship, to determine the speaker's target identity, includes: S601: Based on the target identity of the person being addressed and the measured kinship relationship, query the pre-stored kinship network to determine the speaker's initial identity; S602: Based on the speaker's initial identity, query the pre-stored second mapping table to determine the second basic feature and second frequently used position corresponding to the speaker's initial identity. The second mapping table is used to store the mapping relationship between the user identity and its corresponding basic feature and frequently used position. S603: Based on the second basic feature and second common location corresponding to the speaker's initial identity, verify the speaker's initial identity and determine the speaker's target identity.

[0072] The kinship network refers to the network used to reflect the kinship relationships between users. The speaker's initial identity refers to the user identity of the speaker initially determined based on the kinship network.

[0073] As an example, in step S601, the control device queries the kinship network pre-stored in the system database, which reflects the kinship relationships between user identities, for user identities that have a measured kinship relationship with the target identity of the person being addressed, and determines the retrieved user identity as the initial speaker identity. Taking the user identities in Table 1 as an example, the kinship network includes the father-son relationship between vehicle owner A and user B, the husband-wife relationship between vehicle owner A and user C, the mother-son relationship between vehicle owner A and user D, the mother-son relationship between user B and user C, the mother-in-law / daughter-in-law relationship between user B and user D, and the grandparent-grandchild relationship between user C and user D. For example, when the control device determines that the target identity of the person being addressed is vehicle owner A and the measured kinship relationship is a father-son relationship (or parent-child relationship), it determines user B as the initial speaker identity. When the control device determines that the target identity of the person being addressed is user B and the measured kinship relationship is a parent-child relationship (including father-son and mother-son relationships), it determines both vehicle owner A and user C as the initial speaker identities.

[0074] Here, the second basic feature refers to the basic feature corresponding to the speaker's initial identity in the second mapping table. The second frequently used position refers to the frequently used position corresponding to the speaker's initial identity in the second mapping table.

[0075] As an example, in step S602, the control device determines the basic features corresponding to the speaker's initial identity in the pre-stored second mapping table as the second basic features, and determines the commonly used position corresponding to the speaker's initial identity as the second commonly used position, so as to verify the speaker's initial identity based on the second basic features and the second commonly used position. For example, when the control device determines that the speaker's initial identity is user B, it queries the second mapping table, where the basic features corresponding to user B are "facial feature b" and "weight 15±2KG" and the commonly used position feature is "left rear row". It then determines "facial feature b" and "weight 15±2KG" as the second basic features and "left rear row" as the second commonly used position.

[0076] As an example, in step S603, the control device first obtains the measured weight corresponding to the second commonly used position, and then compares and analyzes the measured weight with the weight range in the second basic feature to determine whether the measured weight is within the weight range in the second basic feature.

[0077] When the control device determines that the measured weight falls within the weight range of the second basic feature, it identifies the speaker's initial identity as the speaker's target identity. Understandably, since the measured weight is obtained at the second frequently used location corresponding to the speaker's initial identity, if the measured weight falls within the weight range of the second basic feature, it indicates that the second frequently used location corresponds to the speaker. Since the user identity corresponding to the second frequently used location is the speaker's initial identity, the user identity corresponding to the speaker is the speaker's initial identity; that is, the speaker's target identity is the speaker's initial identity.

[0078] When the control device determines that the measured weight is not within the weight range of the second basic features, it indicates that the second commonly used position does not correspond to the speaker's initial identity. The device then obtains the measured weight corresponding to the remaining position in the cabin. If there is a case where the measured weight is within the weight range of the second basic features, the device then obtains the measured facial features of the user at the remaining position corresponding to that measured weight. It then determines whether the measured facial features are consistent with the facial features in the second basic features. If they are consistent, the speaker's initial identity is determined as the speaker's target identity. If they are inconsistent, the device determines that the speaker is a new user who does not belong to the second mapping table. A new user identity is set for the speaker, and the obtained measured weight, measured facial features, and position of the speaker are added to the second mapping table along with the new user identity to complete the second mapping table. If the measured weight does not fall within the weight range of the first basic feature, then it is determined that the speaker is not a new user in the second mapping table. A new user identity is set for the speaker. Based on the full-domain speech data, the speaker is located to determine the speaker's location. The measured weight and measured facial features of the speaker's location are obtained. The speaker's measured weight, measured facial features, and location are added to the second mapping table along with the new user identity to complete the second mapping table.

[0079] For example, if the speaker's initial identity is user B, then as shown in Table 2 above, the second most common position is the left rear row, the facial feature in the second basic feature is "facial feature b", and the weight range in the second basic feature is 15±2KG. The control device first obtains the measured weight corresponding to the left rear row, and then compares and analyzes the measured weight with the weight range of 15±2KG in the second basic feature to determine whether the measured weight is within the range corresponding to 15±2KG.

[0080] When the control device determines that the measured weight is within the weight range of 15±2KG in the second basic feature, it indicates that the speaker who issued the address is in the second commonly used position, and determines the speaker's initial identity "User B" as the speaker's target identity.

[0081] When the control device determines that the measured weight is not within the weight range of 15±2KG in the second basic feature, it indicates that the user identity at the second commonly used position is not the speaker's initial identity, and the speaker at the second commonly used position is not the one who issued the address. The device then obtains the measured weight corresponding to the remaining positions in the cabin. If there is a position where the measured weight is within the weight range of 15±2KG in the second basic feature, the device then obtains the measured facial features of the user at that position through the camera device. It then determines whether the measured facial features are consistent with "facial feature b" in the first basic feature. If they are consistent, the speaker's initial identity "User B" is determined as the speaker's target identity. If they are inconsistent, the device determines that the speaker is a new user not in the second mapping table, sets a new user identity for the speaker, and locates the speaker based on the full-domain voice data to determine the speaker's location. The device then obtains the measured weight and measured facial features of the speaker's location and adds the obtained measured weight, measured facial features, and speaker's location to the second mapping table along with the new user identity to complete the second mapping table. If there is no location with a measured weight within the weight range of 15±2KG in the second basic feature, a new user identity is set for the speaker. Based on the full-domain speech data, the speaker is located to determine the speaker's location. The measured weight and measured facial features of the speaker's location are obtained. The speaker's measured weight, measured facial features, and speaker's location are then added to the second mapping table along with the new user identity to complete the second mapping table.

[0082] In this example, if it is determined that there is more than one initial speaker identity corresponding to the speaker, then S601 to S603 are executed repeatedly to verify the identity of each initial speaker identity and determine the target speaker identity corresponding to the speaker.

[0083] In this embodiment, the initial speaker identity is determined based on the target identity of the person being addressed and the measured kinship relationship. Then, according to a pre-stored second mapping table, the initial speaker identity is verified using the second basic feature and second frequently used location corresponding to the initial speaker identity, thus accurately determining the speaker's target identity. This method eliminates the need for user intervention within the cabin, naturally determining the speaker's target identity and improving the user's driving experience and safety.

[0084] In one embodiment, before step S101, i.e. before acquiring full-domain voice data within the cockpit, the vehicle control method further includes: S1011: Obtain historical dialogue data within the vehicle cabin, and based on the historical dialogue data, determine the common locations and basic characteristics of each speaker. S1012: Extract natural address from historical dialogue data to determine the address and user identity of each interlocutor; S1013: Store the same user identity and its corresponding title to obtain a first mapping table, and store the same user identity and its corresponding common locations and basic features to obtain a second mapping table.

[0085] Historical dialogue data refers to conversations between any two individuals collected within a train carriage during a specific historical time period. The historical time period refers to one or more past time periods. The participants in the conversation are the individuals involved.

[0086] As an example, in step S1011, microphones at different locations in the global microphone array collect historical dialogue data between multiple speakers in the vehicle cabin in real time over a historical period, and send this historical dialogue data to the control device. The control device receives the historical dialogue data in real time over the historical period and determines the position of each microphone transmitting the historical dialogue data through sound source localization, thereby determining the cabin position corresponding to each speaker in the historical dialogue data. It also obtains the weight at each position through a seat pressure sensor in the identity verification module and the facial features at each position through a camera device in the identity verification module, using the weight and facial features of the same position determined each time as basic features. The control device obtains the positions and basic features corresponding to multiple speakers within the historical period in the above manner, statistically analyzes positions with the same basic features, and obtains the top N positions with higher probabilities corresponding to the same basic feature. These top N positions with higher probabilities corresponding to the same basic feature are determined as commonly used positions corresponding to the same basic feature. For example, for facial feature a, the basic feature of weight within the range of 70±2KG, there is a 90% probability of being in the driver's seat and a 10% probability of being in the passenger seat. Therefore, the driver's seat and the passenger seat are determined as the common locations corresponding to this basic feature.

[0087] Natural address extraction refers to extracting address terms from historical dialogue data.

[0088] As an example, in step S1012, the control device extracts natural forms of address from the received historical dialogue data, determines the corresponding address words for the speakers, and sets the user identity of the speakers based on their frequently used locations. For example, as shown in Table 1 above, the user identity of the speaker whose frequently used location is the driver's seat is set as car owner A, the user identity of the speaker whose frequently used location is the left rear seat is set as user B, the user identity of the speaker whose frequently used location is the front passenger seat is set as user C, and the user identity of the speaker whose frequently used location is the right rear seat is set as user D. In this example, the control device can also determine the kinship relationship between two user identities based on the address words, forming a kinship network. For example, as shown in Table 1 above, there is a father-son relationship between car owner A and user B, a husband-wife relationship between car owner A and user C, a mother-son relationship between car owner A and user D, a mother-son relationship between user B and user C, a mother-in-law / daughter-in-law relationship between user B and user D, and a grandparent-grandchild relationship between user C and user D.

[0089] As an example, in step S1013, the control device stores the statistically obtained user identities and their corresponding titles to form a first mapping table. For example, as shown in Table 1, the first mapping table is formed by storing the same user identities (vehicle owner A, user B, user C, and user D) and their corresponding titles.

[0090] The control device stores the statistically obtained user identities and their corresponding frequently used locations and basic characteristics, forming a second mapping table that includes the mapping relationship between user identities, basic characteristics, and frequently used locations. For example, as shown in Table 2, the same user identity (vehicle owner A, user B, user C, and user D) and their corresponding user identities, basic characteristics, and frequently used locations are stored to form the second mapping table.

[0091] In this embodiment, common locations, basic features, and natural names are extracted from historical dialogue data to form a first mapping table and a second mapping table. This facilitates natural and seamless identity verification of users in the cabin based on the first and second mapping tables, providing users with accurate and personalized services.

[0092] This application also provides a vehicle control device 70, please refer to... Figure 7 The system includes: a voice acquisition module 710 for acquiring full-domain voice data within the cockpit; a voice recognition module 720 for performing voice recognition on the full-domain voice data to determine the address words in the full-domain voice data and the measured kinship relationship between the speaker and the addressed person; an identity recognition module 730 for performing identity recognition based on the address words and the measured kinship relationship to determine the speaker target identity and the addressed person target identity; and a personalized service execution module 740 for controlling the target components of the vehicle to execute personalized services based on the speaker target identity and the addressed person target identity.

[0093] This application also provides an electronic device 80, please refer to... Figure 8 It includes a memory 810 and a processor 820, wherein the memory 810 is used to store computer programs; and the processor 820 is used to execute the programs stored in the memory 810 to implement the vehicle control method described in any embodiment of this application.

[0094] This application also provides a vehicle that includes the control device described in the above embodiments.

[0095] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the vehicle control method described in any embodiment of this application.

[0096] In this application, "multiple" refers to two or more.

[0097] In this application, unless otherwise expressly defined, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0098] The terms “first,” “second,” “third,” “fourth,” etc., in this application (if present) are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0099] In this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, in this application, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0100] Unless otherwise specified, all steps in this application may be performed sequentially or randomly. For example, if the method includes steps A and B, it means that the method may include steps A and B performed sequentially, or it may include steps B and A performed sequentially. For example, if the method may also include step C, it means that step C may be added to the method in any order. For example, the method may include steps A, B, and C, or it may include steps A, C, and B, or it may include steps C, A, and B, etc.

[0101] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A vehicle control method, characterized in that, include: Acquire full-area voice data within the cockpit; Speech recognition is performed on the full-domain speech data to determine the address words in the full-domain speech data and the measured kinship relationship between the speaker and the person being addressed; Based on the address terms and the measured kinship relationships, identity recognition is performed to determine the speaker's target identity and the person being addressed's target identity. Based on the speaker's target identity and the called person's target identity, the target components of the vehicle are controlled to perform personalized services.

2. The vehicle control method as described in claim 1, characterized in that, The method of identity recognition based on address terms and measured kinship to determine the speaker's target identity and the person being addressed's target identity includes: Based on the address, the identity of the person being addressed is matched to determine the target identity of the person being addressed. Based on the target identity of the person being addressed and the measured kinship relationship, identity recognition is performed to determine the speaker's target identity.

3. The vehicle control method as described in claim 2, characterized in that, The step of matching the identity of the person being addressed based on the address term to determine the target identity of the person being addressed includes: The query is based on the address term and a pre-stored first mapping table, which stores the mapping relationship between address terms and user identities. When no addressing term exists in the first mapping table, the measured basic features of the person being addressed are obtained, and the pre-stored second mapping table is queried based on the measured basic features. The user identity that matches the measured basic features in the second mapping table is determined as the target identity of the person being addressed. The second mapping table is used to store the mapping relationship between user identity and its corresponding basic features and common locations. When a title exists in the first mapping table, the user identity corresponding to the title in the first mapping table is determined as the initial identity of the person being addressed. Based on the initial identity of the person being addressed, the second mapping table is queried to determine the first basic feature corresponding to the initial identity of the person being addressed. Based on the measured basic feature of the person being addressed and the first basic feature, the initial identity of the person being addressed is verified to determine the target identity of the person being addressed.

4. The vehicle control method as described in claim 3, characterized in that, The measured basic feature includes the measured weight, and the first basic feature includes a first weight range; The step of querying the second mapping table based on the initial identity of the person being addressed to determine the first basic feature corresponding to the initial identity of the person being addressed, and performing identity verification on the initial identity of the person being addressed based on the measured basic feature of the person being addressed and the first basic feature to determine the target identity of the person being addressed, includes: Based on the initial identity of the person being addressed, the second mapping table is queried to determine the first weight range and the first frequently used location corresponding to the initial identity of the person being addressed. If the measured weight of the person being addressed at the first commonly used location is within the first weight range, then the initial identity of the person being addressed is determined as the target identity of the person being addressed. If the measured weight of the person being addressed at the first commonly used location is not within the first weight range, and the measured weight of the person being addressed at the remaining locations is within the first weight range, obtain the voice response of the person being addressed. Based on the voice response of the person being addressed, verify the initial identity of the person being addressed and determine the target identity of the person being addressed.

5. The vehicle control method as described in claim 4, characterized in that, The first basic feature also includes a first facial feature; The process of verifying the initial identity of the person being addressed based on their voice response, and determining the target identity of the person being addressed, includes: Based on the voice response of the person being addressed, the location of the person being addressed is determined, and the actual measured location corresponding to the person being addressed is determined. If the measured location is the same as the remaining location, then the initial identity of the person being called is determined as the target identity of the person being called; If the measured location is not the same as the remaining location, the measured facial features corresponding to the measured location are obtained. If the measured facial features are consistent with the first facial features, the initial identity of the person being called is determined as the target identity of the person being called.

6. The vehicle control method as described in claim 2, characterized in that, The step of identifying the speaker's target identity based on the target identity of the person being addressed and the measured kinship relationship includes: Based on the target identity of the person being addressed and the measured kinship, the pre-stored kinship network is queried to determine the initial speaker identity corresponding to the speaker; Based on the speaker's initial identity, a pre-stored second mapping table is queried to determine the second basic feature and second frequently used position corresponding to the speaker's initial identity. The second mapping table is used to store the mapping relationship between the user identity and its corresponding basic feature and frequently used position. Based on the second basic feature and second common location corresponding to the speaker's initial identity, the speaker's initial identity is verified to determine the speaker's target identity.

7. A vehicle control device, characterized in that, include: The voice acquisition module is used to acquire full-area voice data within the cockpit; The speech recognition module is used to perform speech recognition on the full-domain speech data to determine the address words in the full-domain speech data and the measured kinship relationship between the speaker and the person being addressed; The identity recognition module performs identity recognition based on the address and the measured kinship relationship, and determines the speaker target identity corresponding to the speaker and the addressee target identity corresponding to the addressee. The personalized service execution module controls the target components of the vehicle to execute personalized services based on the speaker's target identity and the called person's target identity.

8. A control device, characterized in that, Including processor and memory, among which, Memory, used to store computer programs; A processor for executing a program stored in memory to implement the vehicle control method according to any one of claims 1-6.

9. A vehicle, characterized in that, Includes the control device as described in claim 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the vehicle control method according to any one of claims 1-6.

Images