Vehicle voice interaction method, vehicle and storage medium

Abstract

The present application discloses a vehicle voice interaction method, comprising: receiving the segmentation processing results of a user's voice request sent by a server; confirming the execution instruction type corresponding to each sentence based on the segmentation processing results; if the execution instruction type corresponding to the current sentence is a predetermined type, pre-processing the execution instruction corresponding to the current sentence; if the execution instruction type corresponding to the current sentence is not a predetermined type, immediately issuing the execution instruction corresponding to the current sentence; executing the execution instruction corresponding to each sentence and broadcasting the execution result to complete the voice interaction. In the present application, the voice request is streamed, which speeds up the overall execution speed of the voice interaction. At the same time, when issuing the execution instructions corresponding to each sentence, the execution type corresponding to each sentence is judged, thereby avoiding the possibility of inconsistency between broadcast and execution caused by issuing the execution instructions too quickly, thereby improving the user experience.

Description

Technical Field

[0001] The present application relates to the field of transportation, and in particular to a vehicle voice interaction method, a vehicle, and a computer-readable storage medium. Background Art

[0002] The voice assistant of the in-vehicle system provides many conveniences for users in the cockpit. Users can control the software or vehicle components in the cockpit through voice interaction. In order to provide users with a more convenient voice interaction method, the relevant technology supports the interaction method in which the user's request includes multiple sentences. For example, the voice request issued by the user is "report today's weather, straighten the seat back, and adjust the air conditioner to 26℃". For this type of voice request including multiple sentences, the relevant technology will bind the execution instructions corresponding to each sentence and the feedback broadcast of the execution, and issue them in sequence, executing one broadcast at a time, and then issuing the next one. However, this will result in a slower overall execution speed of the interaction, affecting the user experience. Summary of the Invention

[0003] In view of this, the present application provides a vehicle voice interaction method, a vehicle, and a computer-readable storage medium.

[0004] This application provides a voice interaction method, including:

[0005] receiving a sentence segmentation processing result of a user voice request sent by a server, where the user voice request includes a plurality of sentences;

[0006] Determine the execution instruction type corresponding to each sentence according to the sentence segmentation processing result;

[0007] If the execution instruction type corresponding to the current sentence is a predetermined type, pre-processing the execution instruction corresponding to the current sentence;

[0008] If the execution instruction type corresponding to the current sentence is not a predetermined type, immediately issuing the execution instruction corresponding to the current sentence;

[0009] Execute the execution instruction corresponding to each sentence and broadcast the execution result to complete the voice interaction.

[0010] In this way, the voice interaction method of the present application no longer binds execution and broadcasting for voice requests sent by users that include multiple sentences. Instead, it streams the requests, which speeds up the overall execution speed of the voice interaction. At the same time, the server segments the voice requests and sends the processed results to the vehicle. When sending the execution instructions corresponding to each sentence, the execution type corresponding to each sentence is determined. If the current sentence belongs to a predetermined type, it is pre-processed before being sent. If the current sentence does not belong to a predetermined type, it is sent immediately. This avoids the possibility of inconsistencies between broadcasting and execution caused by sending the execution instructions too quickly, thereby improving the user experience.

[0011] The step of determining the execution instruction type corresponding to each sentence according to the sentence segmentation processing result includes:

[0012] The execution instruction type corresponding to the current sentence is determined according to the intention of the current sentence.

[0013] In this way, the execution instruction type corresponding to the current sentence can be confirmed based on the intention identified by semantic recognition of the current sentence.

[0014] When the execution instruction type corresponding to the current sentence is a predetermined type, pre-processing the execution instruction corresponding to the current sentence includes:

[0015] When the execution instruction type corresponding to the current sentence is a first predetermined type, issuing the execution instruction corresponding to the current sentence;

[0016] Delay processing is performed after the execution instruction corresponding to the current sentence is issued.

[0017] In this way, when it is confirmed that the execution instruction type corresponding to the current sentence is the first predetermined type, the issuance of the execution instruction needs to be preprocessed. Specifically, after the execution instruction corresponding to the current sentence is issued, a certain time is delayed before issuing the subsequent execution instruction.

[0018] The delay processing after issuing the execution instruction corresponding to the current sentence includes:

[0019] After the execution result of the execution instruction corresponding to the current sentence is broadcasted, the execution instruction corresponding to the next sentence adjacent to the current instruction is issued.

[0020] In this way, the time for delaying the issuance of subsequent instructions is the duration of broadcasting the execution result of the execution instruction corresponding to the current sentence. After the execution result of the execution instruction corresponding to the current sentence is broadcast, the execution instruction corresponding to the subsequent sentence is issued, thereby preventing the inconsistency between execution and broadcasting due to the long broadcasting time of the execution result of the execution instruction corresponding to the current sentence.

[0021] The step of determining the execution instruction type corresponding to each sentence according to the sentence segmentation processing result includes:

[0022] Determining the execution instruction type corresponding to the current sentence based on the logical relationship between the current sentence and the sentence before the current sentence;

[0023] In the case that there is a predetermined logical relationship between the current sentence and the sentence before the current sentence, confirm that the instruction type corresponding to the current sentence is a second predetermined type, wherein the predetermined logical relationship includes a logical relationship in which there is a conflicting logical relationship between the execution actions for the same object in the current sentence and the sentence before the current sentence.

[0024] In this way, in addition to the intention of each clause itself, the execution instruction type corresponding to the current clause is also judged based on the logical relationship between the clauses. When there is a predetermined logical relationship, it is confirmed that the current clause belongs to the second predetermined type, and preprocessing is required when issuing the instruction.

[0025] When the execution instruction type corresponding to the current sentence is a predetermined type, pre-processing the execution instruction corresponding to the current sentence includes:

[0026] In a case where the execution instruction type corresponding to the current sentence is a second predetermined type, delay processing is performed before the execution instruction corresponding to the current sentence is issued.

[0027] Thus, for the clauses of the second predetermined type, delayed processing is required when the clauses are issued, so as to prevent the current execution instruction from being unable to execute normally due to conflicting execution logics between the execution instructions corresponding to the preceding and following clauses.

[0028] The delay processing before issuing the execution instruction corresponding to the current sentence includes:

[0029] Before issuing the execution instruction corresponding to the current sentence, a predetermined time is delayed and the execution instruction corresponding to the current sentence is issued.

[0030] Thus, for a clause of the second predetermined type of execution instruction, it is necessary to delay issuing the corresponding execution instruction for a predetermined time before issuing the execution instruction corresponding to the current clause, thereby ensuring the execution success rate of the execution instruction corresponding to the current clause.

[0031] The executing the execution instruction corresponding to each sentence and broadcasting the execution result to complete the voice interaction includes:

[0032] The execution result of each of the execution instructions is fused and reported after the fusion processing.

[0033] In this way, the execution results of the execution instructions are integrated and then broadcast, which improves the reply efficiency and the user experience of the reply content broadcast.

[0034] The present application also provides a vehicle, which includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the above-mentioned voice interaction method is implemented.

[0035] Thus, in this application, for a voice request sent by a user that includes multiple sentences, the execution is no longer bound to the broadcast, but is streamed and sent, which speeds up the overall execution speed of the voice interaction. At the same time, the server segments the voice request and sends the processed results to the vehicle. When sending the execution instructions corresponding to each sentence, the execution type corresponding to each sentence is judged. If the current sentence belongs to the predetermined type, it is pre-processed before being sent. If the current sentence belongs to a non-predetermined type, it is sent immediately, thereby avoiding the possibility of inconsistencies between broadcast and execution caused by sending the execution instructions too quickly, and improving the user experience.

[0036] The present application also provides a computer-readable storage medium. When the computer-readable storage medium stores a computer program, and when the computer program is executed by one or more processors, the above-mentioned voice interaction method is implemented.

[0037] Thus, in this application, for a voice request sent by a user that includes multiple sentences, the execution is no longer bound to the broadcast, but is streamed and sent, which speeds up the overall execution speed of the voice interaction. At the same time, the server segments the voice request and sends the processed results to the vehicle. When sending the execution instructions corresponding to each sentence, the execution type corresponding to each sentence is judged. If the current sentence belongs to the predetermined type, it is pre-processed before being sent. If the current sentence belongs to a non-predetermined type, it is sent immediately, thereby avoiding the possibility of inconsistencies between broadcast and execution caused by sending the execution instructions too quickly, and improving the user experience.

[0038] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become obvious from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0040] Figure 1 This is a flow chart of the vehicle voice interaction method of the present application;

[0041] Figure 2This is a flow chart of the vehicle voice interaction method of the present application;

[0042] Figure 3 This is a flow chart of the vehicle voice interaction method of the present application;

[0043] Figure 4 This is a schematic diagram of the connection status of the computer-readable storage medium and the processor of the interactive method of the present application. DETAILED DESCRIPTION

[0044] The present application is described in detail below. Examples of embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions. The following description with reference to the accompanying drawings is illustrative and is only used to explain the present application and is not to be construed as limiting the present application.

[0045] See also Figure 1 , this application provides a vehicle voice interaction method, comprising the following steps:

[0046] 01: Receive the sentence segmentation processing result of the user's voice request sent by the server;

[0047] 02: Confirm the execution instruction type corresponding to each clause based on the sentence segmentation results;

[0048] 03: When the execution instruction type corresponding to the current clause is a predetermined type, pre-process the execution instruction corresponding to the current clause;

[0049] 04: If the execution instruction type corresponding to the current clause is not a predetermined type, the execution instruction corresponding to the current clause is immediately issued;

[0050] 05: Execute the execution instructions corresponding to each sentence and broadcast the execution results to complete the voice interaction.

[0051] The present application also provides a vehicle, and the voice interaction method of the present application can be implemented by the vehicle of the present application. The vehicle includes a memory and a processor, wherein the memory stores a computer program, and the processor is configured to receive a segmentation processing result of a user's voice request issued by a server, and to determine the execution instruction type corresponding to each sentence based on the segmentation processing result, and to pre-process the execution instruction corresponding to the current sentence if the execution instruction type corresponding to the current sentence is a predetermined type, and to immediately issue the execution instruction corresponding to the current sentence if the execution instruction type corresponding to the current sentence is not a predetermined type, and to execute the execution instruction corresponding to each sentence and broadcast the execution result to complete the voice interaction.

[0052] With the advancement of intelligent vehicle cockpits, the functionality within them is becoming increasingly diverse. For example, display screens are now installed at various locations within the cockpit, such as the passenger and rear seats. Zoned air conditioning is available for the front and rear rows. Each seat is equipped with ventilation and heating. Users can interact with the vehicle through the voice assistant integrated into the vehicle system. Within the cockpit, users can control vehicle components such as windows, seats, and air conditioning through voice interaction, as well as vehicle control, navigation, and audio and video entertainment.

[0053] In order to facilitate users to conduct more convenient voice interaction, the relevant technology supports the interaction method in which a user's voice request includes multiple sentences. For example, the voice request issued by the user is "broadcast today's weather, straighten the seat back, and adjust the air conditioner to 26℃". For this type of voice request including multiple sentences, the relevant technology will bind the execution instructions corresponding to each sentence and the execution feedback broadcast, and issue them in sequence, executing one and broadcasting one, and then issuing the next one. That is, the execution instruction for querying the weather is issued first, and then the queried weather results are broadcast. After the weather broadcast is completed, the execution instruction for seat back adjustment is issued, and then the execution feedback of seat back adjustment is broadcast. After the seat back adjustment execution feedback broadcast is completed, the execution instruction for air conditioning adjustment is issued.

[0054] However, binding the execution instructions with the broadcast and executing them one by one will result in a slower execution speed of the entire interaction process, affecting the user experience.

[0055] In this application, for user voice requests similar to the above, which include multiple clauses, the execution and broadcast are no longer bound together. Instead, a streaming method is used to speed up the overall execution speed of voice interaction. Feedback broadcasts of execution are also delivered in a similar manner. In the above example, "Broadcast today's weather, adjust the seat back, and adjust the air conditioner to 26°C", this application will stream the instructions for broadcasting the weather forecast, adjusting the seat back, and adjusting the air conditioner, and then stream the broadcast content.

[0056] Understandably, with streaming delivery, command execution is no longer tied to the broadcast, and command execution no longer waits for the broadcast to complete. This can lead to inconsistencies between broadcast and execution. For example, because weather broadcasts are typically long and complex, seat back adjustments or even air conditioning adjustments can occur during the weather broadcast. Even after the action is fully executed, the user is still in the process of receiving feedback on the weather or other actions. While this speeds up execution, the significant disconnect between execution and broadcasting can also negatively impact the user experience.

[0057] Therefore, while streaming execution, this application confirms the execution instruction type corresponding to each clause after sentence segmentation. It categorizes execution instructions that may cause abnormalities in execution and broadcast, such as those described above, into a predetermined type. Before issuing the execution instruction corresponding to the corresponding clause, if the execution instruction type corresponding to the current clause is confirmed to be of the predetermined type, its issuance is pre-processed, such as by delaying its issuance. This avoids possible abnormalities, such as inconsistencies between execution and broadcast, and ensures normal voice interaction. For example, in the aforementioned voice request "Report today's weather, straighten the seat backs, and adjust the air conditioner to 26°C," the execution instruction type corresponding to the clause "Report today's weather" is of the predetermined type. Pre-processing is performed before issuing it. For example, considering the length of the broadcast content, the issuance of execution instructions for subsequent clauses can be suspended after the execution instruction is issued, and the weather broadcast can be monitored. Subsequent execution instructions can be issued after the weather broadcast ends. This ensures that execution and broadcast are essentially consistent for the user, resulting in a better user experience in terms of execution speed and feedback consistency.

[0058] Among them, in this application, the processing of user voice requests, that is, sentence segmentation processing, can be performed on the cloud server. In order to improve the processing speed of voice requests, the server can adopt streaming speech recognition (SSR). Streaming speech recognition can be understood as recognition while the user speaks, or it can be understood as recognition while collecting voice signals, rather than recognition after collecting all voice signals, that is, the server receives the voice request sent by the user forwarded by the vehicle in real time, recognizes the currently collected streaming voice, obtains the current recognition text, and then performs N-gram segmentation (N-Gram) on the current recognition text, and inputs the segmentation result into the pre-trained model to predict the probability of the next word appearing after the current recognition text to determine whether to perform sentence segmentation after the current recognition text. Each time a sentence is recognized, it is returned to the vehicle end, and the vehicle end performs subsequent processing. Streaming distribution means that the execution instruction corresponding to each sentence is issued immediately.

[0059] The scheduled types of command execution can be configured before shipment and can be updated later via OTA or other methods. The purpose of setting the scheduled types is to aggregate voice requests with the same execution or broadcast characteristics and perform similar processing, thereby avoiding situations that may affect the user experience.

[0060] The pre-processing of the execution instruction is the processing of the issuance timing, such as whether to issue it immediately or delay it, and how long to delay it.

[0061] If the execution instruction corresponding to the sentence is not of the predetermined type, the execution instruction corresponding to each sentence will be issued immediately based on the processing results of the streaming segmentation, so that the vehicle can quickly respond to the user's voice request.

[0062] In summary, the voice interaction method of the present application no longer binds execution and broadcasting to voice requests sent by users that include multiple sentences. Instead, the voice request is streamed and distributed, which speeds up the overall execution speed of the voice interaction. At the same time, the server segments the voice request and distributes the processed results to the vehicle. When distributing the execution instructions corresponding to each sentence, the execution type corresponding to each sentence is determined. If the current sentence belongs to a predetermined type, it is pre-processed before being distributed. If the current sentence belongs to a non-predetermined type, it is distributed immediately. This avoids the possibility of inconsistencies between broadcasting and execution caused by the execution instructions being distributed too quickly, thereby improving the user experience.

[0063] Step 02 includes:

[0064] 020: Confirm the execution instruction type corresponding to the current clause based on the intention of the current clause.

[0065] The processor is used to determine the execution instruction type corresponding to the current sentence according to the intention of the current sentence.

[0066] Specifically, the execution instruction type corresponding to the current sentence can be confirmed based on the intent recognition results of the sentence. In vehicle usage scenarios, intents may include vehicle control, music, queries, etc. In actual operation, the execution type of the execution instruction can be divided into predetermined type and non-predetermined type based on whether the execution instructions and broadcasts corresponding to different intents will cause abnormal execution scenario status. The predetermined type is the voice request corresponding to the execution instruction that may have abnormal conditions.

[0067] In this way, the execution instruction type corresponding to the current sentence can be confirmed based on the intention identified by semantic recognition of the current sentence.

[0068] See also Figure 2 , step 03 includes:

[0069] 030: When the execution instruction type corresponding to the current clause is the first predetermined type, issue the execution instruction corresponding to the current clause;

[0070] 031: Delay processing after issuing the execution instruction corresponding to the current clause.

[0071] The processor is configured to issue the execution instruction corresponding to the current sentence when the execution instruction type corresponding to the current sentence is a first predetermined type, and to perform delay processing after issuing the execution instruction corresponding to the current sentence.

[0072] Specifically, the first predetermined type is query-type. Query-type instructions are instructions for querying specific content, such as weather, air quality, or vehicle status information, such as remaining battery life or total mileage. Understandably, query-type instructions typically take a longer time to report the execution content. If a subsequent execution instruction is issued immediately after a query-type execution instruction, inconsistencies between the execution and the report may occur.

[0073] For example, if the user's voice request is "report today's weather, open the music page," if no processing is performed, the in-vehicle system's user interface will display the open music application page during the weather report. In this application, considering that the reply content of query-type responses may be long, after issuing the execution command corresponding to the current sentence, the execution command corresponding to the subsequent sentence will be issued after a certain delay.

[0074] In this way, when it is confirmed that the execution instruction type corresponding to the current sentence is the first predetermined type, the issuance of the execution instruction needs to be preprocessed. Specifically, after the execution instruction corresponding to the current sentence is issued, a certain time is delayed before issuing the subsequent execution instruction.

[0075] Step 031 includes:

[0076] 0310: After the execution result of the execution instruction corresponding to the current sentence is broadcast, the execution instruction corresponding to the next sentence adjacent to the current instruction is issued.

[0077] The processor is used to wait for the execution result of the execution instruction corresponding to the current sentence to be broadcasted, and then issue the execution instruction corresponding to the next sentence adjacent to the current instruction.

[0078] Specifically, in this application, the delay time is set to the broadcasting duration of the execution result of the execution instruction corresponding to the current sentence, and the execution instruction corresponding to the next sentence will be issued after the execution result of the current sentence is broadcast. For example, for the above example "broadcast today's weather, open the music page", if it is confirmed that the execution instruction type corresponding to the current sentence "broadcast today's weather" is a query type, the execution instruction corresponding to the sentence "open the music page" will be issued after the weather information is broadcast. This prevents the user interface of the vehicle system from displaying other content during the weather information broadcast.

[0079] For another example, the user's voice request is "report today's weather, open the car window". If no processing is done, the car window will be opened when the weather voice is broadcast, and the voice feedback of the car window opening will not be broadcast until the weather broadcast is completed. In this application, when it is confirmed that the execution instruction type corresponding to the current sentence "report today's weather" is a query type, the execution instruction corresponding to the "open the car window" sentence will be issued after the weather information is broadcast. Avoid the situation where the car window is opened during the weather information broadcast, and the feedback broadcast of the car window opening and the execution of the car window opening are too long, causing a sense of disconnection in the user experience. The issuance of the next sentence will still first confirm the corresponding instruction type to confirm whether preprocessing is required.

[0080] In this way, the time for delaying the issuance of subsequent instructions is the duration of broadcasting the execution result of the execution instruction corresponding to the current sentence. After the execution result of the execution instruction corresponding to the current sentence is broadcast, the execution instruction corresponding to the subsequent sentence is issued, thereby preventing the inconsistency between execution and broadcasting due to the long broadcasting time of the execution result of the execution instruction corresponding to the current sentence.

[0081] See also Figure 3 , step 02 includes:

[0082] 021: Determine the execution instruction type corresponding to the current clause based on the logical relationship between the current clause and the clause before the current clause;

[0083] 022: When a predetermined logical relationship exists between the current clause and the clause before the current clause, confirm that the instruction type corresponding to the current clause is a second predetermined type.

[0084] The processor is used to confirm the execution instruction type corresponding to the current sentence based on the current sentence and the logical relationship between the sentence and the sentence before the current sentence, and to confirm that the instruction type corresponding to the current sentence is a second predetermined type when there is a predetermined logical relationship between the current sentence and the sentence before the current sentence.

[0085] Specifically, the predetermined logical relationship includes a conflicting logical relationship between the execution actions for the same object in the current sentence and the sentence before the current sentence. When such a logical relationship exists between the sentences, it can be considered that the execution instruction corresponding to the current sentence belongs to the second predetermined type. The second predetermined type is a conflicting instruction. In the related art, for this type of instruction, due to the logical relationship between it and the upper and lower clauses, when the vehicle executes this type of execution instruction, the execution instruction cannot be executed normally based on the protection strategy of the electronic components. For example, the voice request is "open the window, close the window", and the current sentence is "close the window". Opening the window and closing the window are completely opposite operations performed on the window. Therefore, if the issuance of such instructions is not processed, the vehicle will no longer execute the closing window after executing the execution instruction to open the window.

[0086] Understandably, these voice requests for executing commands may be caused by the vehicle's location changing within a very short period of time, leading the user to issue consecutive voice requests for the opposite action, or by the user correcting a previous, incorrectly issued voice request. In this case, the later sentence may be the one the user prefers. Therefore, to ensure the successful execution of these commands, certain processing is required.

[0087] In this way, in addition to the intention of each clause itself, the execution instruction type corresponding to the current clause is also judged based on the logical relationship between the clauses. When there is a predetermined logical relationship, it is confirmed that the current clause belongs to the second predetermined type, and preprocessing is required when issuing the instruction.

[0088] Please refer again Figure 3 , step 03 includes:

[0089] 032: When the execution instruction type corresponding to the current clause is the second predetermined type, delay processing is performed before issuing the execution instruction corresponding to the current clause.

[0090] The processor is configured to perform delay processing before issuing the execution instruction corresponding to the current sentence when the execution instruction type corresponding to the current sentence is a second predetermined type.

[0091] Specifically, as mentioned above, for execution instructions that are issued consecutively within a very short period of time and have opposite execution actions, for protection purposes, the current execution instruction will not be executed after the previous execution instruction is executed. In this application, when it is determined that the execution instruction type corresponding to the current clause belongs to the second predetermined type, a certain delay will be performed before issuing the execution instruction corresponding to the current clause. After the delay, the execution instruction corresponding to the current clause will be issued again, thereby ensuring that the execution instruction of the current clause can be executed normally.

[0092] For example, in the example above, "Open window, close window," the current clause "Close window" is logically in conflict with the previous clause "Open window." The execution instruction type corresponding to the current clause "Close window" is determined to be the second predetermined type, i.e., a conflicting instruction. Without any action, the vehicle system would only execute the instruction to open the window, not the instruction to close it. However, in this application, a delay is added before issuing the "Close window" instruction to ensure that the instruction can be executed normally.

[0093] For example, if the user voice request is "Change the display mode to day, change the display mode to night", if no processing is done, the vehicle system will only execute the execution instruction of "Change the display mode to day" because the UI rendering requires a certain amount of time. In this application, if the execution instruction type corresponding to the current sentence "Change the display mode to night" is confirmed to be a conflicting instruction, the execution instruction corresponding to "Change the display mode to night" will be delayed for a certain period of time before being issued, so as to ensure that there is sufficient rendering time, and the rendering of the night display mode can be performed after the rendering of the day display mode.

[0094] Thus, for the clauses of the second predetermined type, delayed processing is required when the clauses are issued, so as to prevent the current execution instruction from being unable to execute normally due to conflicting execution logics between the execution instructions corresponding to the preceding and following clauses.

[0095] Step 032 includes:

[0096] 0320: Delay for a predetermined time before issuing the execution instruction corresponding to the current clause, and then issue the execution instruction corresponding to the current clause.

[0097] The processor is used to delay for a predetermined time before issuing the execution instruction corresponding to the current sentence, and then issue the execution instruction corresponding to the current sentence.

[0098] Specifically, the delay duration of the second predetermined type of instruction can be determined based on vehicle-related information such as the protection duration during vehicle electrical design and UI rendering duration in scenarios where conflicting instructions may exist, or it can be set to a fixed duration, such as 200ms, 500ms, etc.

[0099] Thus, for a clause of the second predetermined type of execution instruction, it is necessary to delay issuing the corresponding execution instruction for a predetermined time before issuing the execution instruction corresponding to the current clause, thereby ensuring the execution success rate of the execution instruction corresponding to the current clause.

[0100] The subsequent delivery of the clause will still first confirm the corresponding instruction type to determine whether preprocessing is required.

[0101] Step 04 includes:

[0102] 040: The execution results of each execution instruction are integrated and reported after integration.

[0103] The processor is used to perform fusion processing on the execution result of each execution instruction and broadcast it after the fusion processing.

[0104] Specifically, we consider fusing the voice broadcasts of all executed actions at the same time. Since the fusion itself requires a certain amount of processing time, the broadcast will start after the fusion is completed. There will be a certain time difference between the broadcast and the action execution. This can basically ensure that the broadcast forms a more appropriate connection after the action is executed, so that users can quickly obtain feedback.

[0105] After the execution instructions corresponding to all clauses are issued, the voice interaction process ends, or if the voice assistant of the vehicle system does not receive a new voice instruction within a predetermined time, such as 1000ms, the voice interaction process ends.

[0106] In this way, the execution results of the execution instructions are integrated and then broadcast, which improves the reply efficiency and the user experience of the reply content broadcast.

[0107] See also Figure 4 The present application also provides a non-volatile computer-readable storage medium 100 including a computer program 101. When the computer program 101 is executed by one or more processors 200, the one or more processors 200 execute the voice interaction method of any of the above embodiments.

[0108] The embodiment of the present invention also provides a computer-readable storage medium. One or more non-volatile computer-readable storage media storing a computer program, when the computer program is executed by one or more processors, implement the interactive method of any of the above-mentioned embodiments. Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant software through a computer program. The program can be stored in a non-volatile computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), etc.

[0109] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with an embodiment or example is included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0110] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable actions for implementing a specific logical function or process step, and the scope of the preferred embodiments of the present application includes additional implementations in which functions may be performed in a different order than shown or discussed, including performing functions in a substantially simultaneous manner or in a reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0111] Although the present application has been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A vehicle voice interaction method, characterized in that: include: receiving a sentence segmentation processing result of a user voice request sent by a server, where the user voice request includes a plurality of sentences; Determine the execution instruction type corresponding to each sentence according to the sentence segmentation processing result; When the execution instruction type corresponding to the current sentence is a first predetermined type, issuing the execution instruction corresponding to the current sentence, and performing delay processing after issuing the execution instruction corresponding to the current sentence; If the execution instruction type corresponding to the current sentence is not a predetermined type, immediately issuing the execution instruction corresponding to the current sentence; Execute the execution instruction corresponding to each sentence and broadcast the execution result to complete the voice interaction.

2. The method according to claim 1, characterized in that The step of determining the execution instruction type corresponding to each sentence according to the sentence segmentation processing result includes: The execution instruction type corresponding to the current sentence is determined according to the intention of the current sentence.

3. The method according to claim 1, characterized in that The delay processing after issuing the execution instruction corresponding to the current sentence includes: After the execution result of the execution instruction corresponding to the current sentence is broadcasted, the execution instruction corresponding to the next sentence adjacent to the current instruction is issued.

4. The method according to claim 1, wherein The step of determining the execution instruction type corresponding to each sentence according to the sentence segmentation processing result includes: Determining the execution instruction type corresponding to the current sentence based on the logical relationship between the current sentence and the sentence before the current sentence; In the case that there is a predetermined logical relationship between the current sentence and the sentence before the current sentence, confirm that the instruction type corresponding to the current sentence is a second predetermined type, wherein the predetermined logical relationship includes a logical relationship in which there is a conflicting logical relationship between the execution actions for the same object in the current sentence and the sentence before the current sentence.

5. The method according to claim 1, wherein The method further comprises: In a case where the execution instruction type corresponding to the current sentence is a second predetermined type, delay processing is performed before the execution instruction corresponding to the current sentence is issued.

6. The method according to claim 5, characterized in that The delay processing before issuing the execution instruction corresponding to the current sentence includes: Before issuing the execution instruction corresponding to the current sentence, a predetermined time is delayed and the execution instruction corresponding to the current sentence is issued.

7. The method according to claim 1, characterized in that The executing the execution instruction corresponding to each sentence and broadcasting the execution result to complete the voice interaction includes: The execution result of each of the execution instructions is fused and reported after the fusion processing.

8. A vehicle, characterized in that: The vehicle includes a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the method according to any one of claims 1 to 7 is implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by one or more processors, the method according to any one of claims 1 to 7 is implemented.

Images