A data processing method, device, medium and product

Abstract

The application discloses a data processing method and device, medium and product. It relates to the technical field of vehicles, and comprises the following steps: acquiring input information; determining a target interaction scene according to the input information; determining a target confidence of target vehicle memory data in the target interaction scene according to the spatial confidence, the time confidence, the semantic confidence, the context confidence of the target vehicle memory data corresponding to the input information and the multidimensional weight of the target interaction scene; inputting the target confidence of the target vehicle memory data in the target interaction scene, the input information and the target vehicle memory data into a target model to obtain a target control instruction. The technical scheme of the application can improve the accuracy of the determined control instruction.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of vehicle technology, and in particular to a data processing method, device, medium and product. Background Technology

[0002] With the continuous evolution of intelligent connected vehicles and vehicle domain control systems, in-vehicle systems have gradually acquired long-term memory capabilities, enabling them to continuously record and store heterogeneous data from multiple sources.

[0003] Because sensor data in multi-source heterogeneous vehicle memory data is prone to spatiotemporal conflicts or semantic contradictions, directly inputting the above vehicle memory data into the target model can easily result in incorrect control commands. Summary of the Invention

[0004] This invention provides a data processing method, device, medium, and product that can solve the above-mentioned problems and improve the accuracy of determined control commands.

[0005] According to one aspect of the present invention, a data processing method is provided, comprising: Obtain input information; Based on the input information, determine the target interaction scenario; Based on the spatial confidence, temporal confidence, semantic confidence, contextual confidence of the target vehicle memory data corresponding to the input information, and the multi-dimensional weights of the target interaction scenario, the target confidence of the target vehicle memory data under the target interaction scenario is determined. In the target interaction scenario, the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data are input into the target model to obtain the target control command.

[0006] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and a memory communicatively connected to said at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the data processing method according to any embodiment of the present invention.

[0007] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the data processing method described in any embodiment of the present invention.

[0008] According to another aspect of the present invention, a computer program product is provided, which, when executed by a processor, implements the data processing method as described in any of the embodiments of the present invention.

[0009] This invention, through the following embodiments, acquires input information; determines a target interaction scenario based on the input information; determines the target confidence level of the target vehicle memory data under the target interaction scenario based on the spatial confidence level, temporal confidence level, semantic confidence level, contextual confidence level of the target vehicle memory data corresponding to the input information, and the multi-dimensional weights of the target interaction scenario; inputs the target confidence level of the target vehicle memory data under the target interaction scenario, the input information, and the target vehicle memory data into a target model to obtain target control commands. This allows for the pre-determination of the target confidence level of the target vehicle memory data, and by inputting the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data into the target model to obtain target control commands, resulting in more accurate control commands.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of a data processing method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0014] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0015] It is understood that before using the technical solutions disclosed in the various embodiments of this disclosure, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in this disclosure in an appropriate manner in accordance with relevant laws and regulations, and user authorization should be obtained.

[0016] Example 1 Figure 1 This is a flowchart illustrating a data processing method provided in an embodiment of the present invention. This embodiment is applicable to data processing situations. The method can be executed by the data processing device in this embodiment of the present invention, which can be implemented in software and / or hardware, such as... Figure 1 As shown, the method specifically includes the following steps: S110, Obtain input information.

[0017] In this embodiment, the input information can be user-inputted information, used in vehicle-to-everything (V2X) interaction scenarios. The input information can be voice signals, touch information, text, gesture information, etc.

[0018] S120, determine the target interaction scenario based on the input information.

[0019] In this embodiment, the method for determining the target interaction scenario based on the input information can be as follows: the input information is segmented into frames, acoustic features are extracted, and the acoustic features are input into a classifier to obtain the interaction scenario type. The target interaction scenario can be any one of a multimedia control scenario, a vehicle body control scenario, or a driver assistance scenario.

[0020] S130, based on the spatial confidence, temporal confidence, semantic confidence, contextual confidence of the target vehicle memory data corresponding to the input information and the multi-dimensional weights of the target interaction scenario, determine the target confidence of the target vehicle memory data under the target interaction scenario.

[0021] In this embodiment, the target vehicle memory data is acquired before the input information is received. The target vehicle memory data includes multi-dimensional data, such as GPS message data, CAN bus data, visual perception data, and voice data. The CAN bus data includes valid data such as engine speed, voltage, vehicle speed, and fault information.

[0022] In this embodiment, the target vehicle memory data can be obtained through GPS messages, CAN bus frames, visual perception results, and voice streams.

[0023] It should be noted that the acquired sensor data can also be pre-packaged into standardized vehicle memory data. Vehicle memory data includes the following fields: raw data field, timestamp field, spatial location field, semantic tag field, context tag field, and confidence field. The confidence field can be implemented using the method provided in this embodiment to determine the target confidence level of the target vehicle memory data in the target interaction scenario. The raw data field includes the raw physical data parsed via the vehicle Ethernet or CAN FD protocol. The time field includes microsecond-level high-precision time information based on vehicle second pulse synchronization, ensuring strict alignment of each sensor's data on the time axis. The spatial location field contains the three-dimensional coordinates, accuracy factor, and positioning mode identifier in the vehicle coordinate system, defining the spatial reference system of the data. This field is filled in after time alignment and coordinate transformation between the raw data field and the positioning data field, indicating the vehicle's position at the time of memory entry retrieval, obtained from positioning sources such as GPS, visual SLAM, and map matching. The semantic tag field stores structured event descriptions generated based on vehicle status data, visual perception results, voice interaction results, and map information from the raw data field. The system reads the structured output results and extracts the event subject, event action, associated objects, object states, and logical relationships according to a preset vehicle event template, forming initial semantic label fields. Contextual label fields are used to characterize the current driving mode, road scene, and sensor degradation status, providing environmental constraints for subsequent reliability assessment of memory.

[0024] Optionally, based on the spatial confidence, temporal confidence, semantic confidence, contextual confidence, and multi-dimensional weights of the target vehicle memory data corresponding to the input information, the target confidence of the target vehicle memory data under the target interaction scenario is determined, including: Obtain the spatial confidence, temporal confidence, semantic confidence, and contextual confidence of the target vehicle memory data corresponding to the input information.

[0025] In this embodiment, the spatial confidence, temporal confidence, semantic confidence, and contextual confidence of the target vehicle memory data corresponding to the input information can be obtained as follows: The target vehicle memory data includes sensor data within a target time period, which is prior to the reception time of the input information; the spatial confidence of the target vehicle memory data is determined based on the physical deviation of the target vehicle memory data, the initial weight of the positioning source corresponding to the target vehicle memory data, and the spatial coordinates of the positioning source, wherein the positioning source corresponding to the target vehicle memory data is the location-related sensor data in the target vehicle memory data; the semantic confidence of the target vehicle memory data is determined based on the target vehicle event type corresponding to the sensor data and the semantic association graph corresponding to the target vehicle event type; the contextual confidence of the target vehicle memory data is determined based on the sensor data corresponding to the environmental perception channel in each environmental dimension of the target vehicle memory data; and the temporal confidence of the target vehicle memory data is determined based on the target vehicle speed, the target adaptive coefficient, and the vehicle speed within the target window corresponding to the target vehicle memory data.

[0026] The multi-dimensional weights of the target interaction scenario are obtained by querying the preset mapping table.

[0027] The multi-dimensional weights include spatial weights, temporal weights, semantic weights, and contextual weights.

[0028] In this embodiment, the preset mapping table includes: multi-dimensional weights corresponding to each interaction scenario.

[0029] The sum of the product of the spatial weight and the spatial confidence, the product of the temporal weight and the temporal confidence, the product of the semantic weight and the semantic confidence, and the product of the contextual weight and the contextual confidence is taken as the target confidence of the target vehicle memory data in the target interaction scenario.

[0030] In this embodiment, the target confidence level of the target vehicle memory data under the target interaction scenario is determined based on the following formula: ; in, The target confidence level of the target vehicle's memory data in the target interaction scenario. Spatial weights, For spatial confidence, As time weight, For time confidence, For semantic weights, For semantic confidence, For context weights, Contextual confidence level.

[0031] It should be noted that multi-dimensional weights can be dynamically selected based on the specific application scenario to generate a target confidence level for that scenario. These application scenarios include, but are not limited to, navigation route planning, voice interaction content recall, valet parking location retrieval, driving behavior retrospective analysis, and accident liability determination. For different scenarios, corresponding multi-dimensional weights are loaded from a preset scenario-weight mapping table. For example, in the voice interaction recall scenario, semantic confidence has the highest weight (e.g., ...). =0.1, =0.2, =0.5, =0.2).

[0032] Optionally, obtaining the spatial confidence, temporal confidence, semantic confidence, and contextual confidence of the target vehicle memory data corresponding to the input information includes: Obtain the target vehicle memory data corresponding to the input information.

[0033] The target vehicle memory data includes sensor data within a target time period, which is prior to the time the input information is received.

[0034] The spatial confidence level of the target vehicle memory data is determined based on the physical deviation of the target vehicle memory data, the initial weight of the positioning source corresponding to the target vehicle memory data, and the spatial coordinates of the positioning source.

[0035] The location source corresponding to the target vehicle memory data is the location-related sensor data in the target vehicle memory data.

[0036] In this embodiment, determining the physical deviation of the target vehicle's memory data can prevent abnormal data from contaminating subsequent fusion processes.

[0037] In this embodiment, by comparing multiple location sources laterally in the spatial dimension, the system is equipped with the ability to identify conflicts in outlier data based on multi-source consensus.

[0038] Based on the target vehicle event type corresponding to the sensor data and the semantic association graph corresponding to the target vehicle event type, the semantic confidence of the target vehicle memory data is determined.

[0039] The contextual confidence level of the target vehicle's memory data is determined based on the sensor data corresponding to the environmental perception channels in each environmental dimension of the target vehicle's memory data.

[0040] The temporal confidence level of the target vehicle memory data is determined based on the target vehicle speed, target adaptive coefficient, and vehicle speed within the target window corresponding to the target vehicle memory data.

[0041] In this embodiment, the temporal confidence of the target vehicle memory data includes confidence in four independent dimensions: space, time, semantics, and context. This enables downstream applications to obtain the required confidence information for specific dimensions, avoiding misjudgment and blocking of high-quality information due to dimensional information compression.

[0042] Optionally, the spatial confidence level of the target vehicle memory data is determined based on the physical deviation of the target vehicle memory data, the initial weight of the positioning source corresponding to the target vehicle memory data, and the spatial coordinates of the positioning source, including: The target weight of the positioning source corresponding to the target vehicle memory data is determined based on the physical deviation of the target vehicle memory data and the initial weight of the positioning source corresponding to the target vehicle memory data.

[0043] In this embodiment, the physical deviation of the target vehicle memory data can be obtained by processing the motion-related sensor data in the target vehicle memory data using the basic constraints of vehicle kinematics.

[0044] The location source corresponding to the target vehicle memory data is the location-related sensor data in the target vehicle memory data.

[0045] In this embodiment, assuming there are n location sources at the same time t, the weighted consensus position of the j-th location source is determined based on the following formula: ; in, Let j be the weighted consensus position of the j-th location source. Let the initial weights be those of the j-th location source. The physical deviation of the target vehicle's memory data is λ, which is the attenuation coefficient. λ∈[1.0,5.0], and can be preset in the vehicle system according to the sensor calibration parameters.

[0046] In this embodiment, the initial weight of the j-th positioning source is determined by the inherent accuracy of the positioning source.

[0047] Based on the target weight of the location source corresponding to the target vehicle memory data and the spatial coordinates of the location source, the weighted consensus position of the location source is determined.

[0048] In this embodiment, the weighted consensus position of the j-th location source is determined based on the following formula: ; in, Let j be the weighted consensus position of the j-th location source. The vehicle position vector is obtained by time synchronization and coordinate system transformation of the spatial coordinates of the j-th positioning source.

[0049] In this embodiment, the positioning sources include GPS positioning sources, visual SLAM positioning sources, and map matching positioning sources. For positioning sources that output latitude and longitude, they are converted to a unified geographic coordinate system; for visual SLAM positioning sources that output local pose, they are transformed to the same coordinate system through coordinate transformation relationships; for map matching positioning sources, the longitudinal position and lateral offset of the road are converted into unified spatial coordinates. The positioning accuracy and error parameters of different positioning sources are not included in the position vector. In the middle, the original weights are used to determine the corresponding location sources. .

[0050] The consensus deviation of the positioning source is determined based on the spatial coordinates of the positioning source and the weighted consensus position of the positioning source.

[0051] In this embodiment, the consensus bias of the j-th location source is determined based on the following formula: ; in, Let be the consensus deviation of the j-th location source.

[0052] Based on the standard deviation of the consensus bias of the location source and the preset bias threshold, the spatial consensus score of the target vehicle's memory data is determined.

[0053] In this embodiment, the spatial consensus score of the target vehicle's memory data is determined based on the following formula: ; in, Spatial consensus score for the target vehicle's memory data. The standard deviation of the consensus bias of each location source. The maximum permissible deviation is preset (recommended value for vehicle-mounted scenarios). (∈[10m,50m], dynamically adjusted according to road type, with smaller values ​​for highways and larger values ​​for urban canyons).

[0054] It should be noted that when At that time, the spatial consensus score of the target vehicle's memory data is determined to be 0. That is, the spatial consensus score is greater than or equal to 0.

[0055] The spatial confidence level of the target vehicle's memory data is determined based on the spatial consensus score, physical deviation, and preset positioning accuracy parameters of the target vehicle's memory data.

[0056] In this embodiment, the spatial confidence level of the target vehicle's memory data is determined based on the following formula: ; in, Spatial confidence of the target vehicle's memory data. , and Let be the weights of each item, and let the sum of the weights be 1. The preset positioning accuracy parameters are the inherent accuracy parameters of the positioning source.

[0057] Optionally, the target vehicle memory data includes: at least two sensor data corresponding to the target vehicle event type.

[0058] Based on the target vehicle event type corresponding to the sensor data and the semantic association graph corresponding to the target vehicle event type, the semantic confidence of the target vehicle's memory data is determined, including: Determine the set of sensor data pairs based on at least two sensor data points corresponding to the target vehicle event type.

[0059] In this embodiment, any two sensor data points are extracted from at least two sensor data points corresponding to the target vehicle event type to form a sensor data pair. The sensor data is semantically relevant sensor data.

[0060] Based on the semantic association graphs corresponding to the sensor data pairs in the sensor data pair set and the target vehicle event type, determine the number of union nodes, the number of intersection nodes, and the edge weight of the target node for each sensor data pair.

[0061] In this embodiment, the semantic association graph corresponding to the target vehicle event type can be a pre-defined graph.

[0062] The semantic score of the sensor data pair is determined based on the number of union nodes, the number of intersection nodes, and the edge weights of the intersection nodes.

[0063] In this embodiment, the semantic score of a sensor data pair is determined based on the number of union nodes, the number of intersection nodes, and the edge weights of the intersection nodes. The method is as follows: if there is one intersection node, the product of the ratio of the number of intersection nodes to the number of union nodes and the edge weights of the intersection nodes is used as the semantic score of the sensor data pair. If there are multiple intersection nodes, the average edge weights of the multiple intersection nodes are obtained, and the product of the ratio of the number of intersection nodes to the number of union nodes and the average edge weights of the multiple intersection nodes is used as the semantic score of the sensor data pair.

[0064] In this embodiment, the sensor data pair includes: sensor data And sensor data b, the sensor data is determined based on the following formula. Semantic score of sensor data b: ; in, For sensor data The number of nodes intersecting with sensor data b, which is the number of nodes in the sensor data. The number of semantic nodes shared by sensor data b in the semantic association graph corresponding to the target vehicle event type, the For sensor data The number of nodes in the union of sensor data b. For sensor data The edge weights of the nodes intersecting with sensor data b.

[0065] The semantic confidence of the target vehicle memory data is determined based on the semantic score of the sensor data pair and the amount of sensor data corresponding to the target vehicle event type.

[0066] In this embodiment, the target vehicle memory data includes m sensor data corresponding to the target vehicle event type, for example, ,in, This refers to sensor data related to the target vehicle's onboard event type from the target vehicle's memory data. m is greater than or equal to 2.

[0067] In this embodiment, the semantic confidence of the target vehicle's memory data is determined based on the following formula: .

[0068] It should be noted that if the sensor data corresponding to the target vehicle event type has a high degree of coupling in the semantic association graph, then Higher; if the sensor data corresponding to the target vehicle event type contains logical contradictions (such as "green light" and "emergency braking" existing simultaneously without a reasonable explanation path), then Lower.

[0069] Optionally, based on the target vehicle speed, target adaptive coefficient, and vehicle speed within the target window corresponding to the target vehicle's memory data, the temporal confidence level of the target vehicle's memory data is determined, including: The target window length is determined based on the target vehicle speed corresponding to the target vehicle's memory data.

[0070] In this embodiment, the target window length can be dynamically adjusted according to the vehicle speed. A shorter window is used when driving at high speed to ensure real-time performance, while a longer window is used when driving at low speed or when the vehicle is stopped to enhance the smoothing effect.

[0071] In this embodiment, the target window length is determined based on the following formula: ; in, The speed (in m / s) of the target vehicle in the memory data closest to the time of receiving the input information. α is the window adjustment coefficient (α∈[50,200], in meters). To prevent division by zero constant (e.g., ), Minimum window length, This is the maximum window length. The window length is shorter at high speeds to ensure real-time performance, and longer at low speeds or when stationary to enhance smoothing. The target window length.

[0072] Based on the target window length, obtain the vehicle speed, the mean vehicle speed, and the standard deviation of the vehicle speed within the target window.

[0073] In this embodiment, taking the target vehicle's memory data as the center, a certain length of vehicle speed is taken forward and backward to obtain a length of... The vehicle speed within the window.

[0074] The target adaptive coefficient is determined based on the spatial consensus score of the target vehicle's memory data, the initial adaptive coefficient, and the adjustment increment.

[0075] In this embodiment, the target adaptive coefficient is determined based on the following formula: ; in, These are the initial adaptive coefficients. Spatial consensus score for the target vehicle's memory data. To adjust the increment, The target adaptive coefficient.

[0076] If the absolute value of the difference between the vehicle speed in the target window and the mean vehicle speed in the target window is greater than the product of the target adaptive coefficient and the standard deviation of the vehicle speed in the target window, then the first value will be used as the time confidence of the target vehicle memory data.

[0077] In this embodiment, if Then the time confidence level of the target vehicle's memory data is determined to be 1. Wherein, The vehicle speed within the target window. The average vehicle speed within the target window. The standard deviation of vehicle speed within the target window.

[0078] If the absolute value of the difference between the vehicle speed in the target window and the mean vehicle speed in the target window is less than or equal to the product of the target adaptive coefficient and the standard deviation of the vehicle speed in the target window, then the time confidence level of the target vehicle memory data is determined based on the target adaptive coefficient, the vehicle speed in the target window, and the adjustment parameters.

[0079] In this embodiment, the time confidence level of the target vehicle's memory data can be determined based on the target adaptive coefficient, the vehicle speed within the target window, and the adjustment parameters, using the following formula: ; in, The time confidence level of the target vehicle's memory data. For the target adaptive coefficient, The vehicle speed within the target window. The average vehicle speed within the target window. The standard deviation of vehicle speed within the target window. To adjust the parameters.

[0080] In this embodiment, if Then, the temporal confidence level of the target vehicle's memory data is determined based on the above formula.

[0081] Optionally, based on the sensor data corresponding to the environmental perception channels in each environmental dimension of the target vehicle's memory data, the contextual confidence of the target vehicle's memory data is determined, including: The contextual confidence level of the target vehicle's memory data is determined based on the total number of environmental dimensions of the target vehicle's memory data, the total number of environmental perception channels in each environmental dimension, the sensor data corresponding to each environmental perception channel, and the preset weights of each environmental dimension.

[0082] In this embodiment, the contextual confidence level of the target vehicle's memory data can be determined based on the total number of environmental dimensions of the target vehicle's memory data, the total number of environmental perception channels in each environmental dimension, the sensor data corresponding to each environmental perception channel, and the preset weights of each environmental dimension. This can be achieved by using the following formula: ; in, The contextual confidence score of the target vehicle's memory data, where K is the total number of contextual dimensions. Let be the total number of environmental perception channels in the c-th environmental dimension. For the c-th environmental dimension, the first Sensor data corresponding to each environmental perception channel. For the c-th environmental dimension, the first Sensor data corresponding to each environmental perception channel. The preset weights for the c-th environmental dimension.

[0083] In this embodiment, the environmental dimensions include: light, precipitation, temperature, and visibility.

[0084] In this embodiment, the sensor data collected by the camera's illumination detection sensor, the rain sensor, the temperature sensor, and the map weather service, etc., in the target vehicle's memory data are grouped according to environmental dimensions such as illumination, precipitation, temperature, and visibility, and uniformly converted into standard values ​​within the range of [0,1]. .

[0085] It should be noted that, The closer the value is to 1, the more consistent the judgments of multiple environmental perception channels. Only environmental dimensions containing two or more valid channels are included in the calculation.

[0086] S140, the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data in the target interaction scenario are input into the target model to obtain the target control command.

[0087] In this embodiment, the method of inputting the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data into the target model under the target interaction scenario to obtain the target control command can be as follows: if the target confidence level of the target vehicle memory data under the target interaction scenario is greater than or equal to a first confidence threshold, then the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data under the target interaction scenario are input into the target model to obtain the target control command.

[0088] Optionally, the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data in the target interaction scenario are input into the target model to obtain target control commands, including: The timestamp, spatial location vector, voice tag, and context tag of the target vehicle's memory data are obtained.

[0089] In this embodiment, the timestamp is used to record microsecond-level high-precision time information based on vehicle-mounted second pulse synchronization, ensuring strict alignment of sensor data on the time axis. The spatial position vector includes three-dimensional coordinates, precision factor, and positioning mode identifier in the vehicle coordinate system. The spatial position vector is used to define the spatial reference system of the data. After completing time alignment and coordinate transformation, the spatial position vector is determined, which is the position of the vehicle itself, obtained from positioning sources such as GPS, visual SLAM, and map matching. The semantic tag stores descriptive information of structured events, covering driving behavior semantics, associated entity information, and logical causal chains. The semantic tag is used to store structured event description information generated based on vehicle state data, visual perception results, voice interaction results, and map information in the target vehicle's memory data. The output structured results are read, and according to the preset vehicle event template, the event subject, event action, associated object, object state, and logical relationship are extracted to form the initial semantic tag. The context tag is used to characterize the current driving mode, road scene, and sensor condition degradation state.

[0090] The input information, the target vehicle memory data, the timestamp of the target vehicle memory data, the spatial location vector, the voice tag, the context tag, and the target confidence level of the target vehicle memory data under the target interaction scenario are input into the target model to obtain the target control command.

[0091] It should be noted that if the target confidence level of the target vehicle memory data is greater than or equal to the second confidence threshold and less than the first confidence threshold, then the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data in the target interaction scenario are input into the target model to obtain target control commands and confidence level prompts. The confidence level prompts include: a comprehensive confidence level value and an indication of the dimension with the lowest confidence level (e.g., "spatial location may be deviated"), for the user or downstream system to make its own judgment. The first confidence threshold is greater than the second confidence threshold. If the target confidence level of the target vehicle memory data is less than the second confidence threshold, it indicates that the target vehicle memory data is not reliable enough and cannot be used as input to the target model. A confirmation request needs to be actively initiated to the user, who then decides whether to use it. The first and second confidence thresholds can be configured independently according to different application scenarios. Safety-sensitive scenarios (e.g., autonomous driving, valet parking) use higher threshold standards, while information display scenarios (e.g., driving log browsing) use more lenient threshold standards.

[0092] In this embodiment, after executing the target control command, based on the actual feedback value The difference between the target confidence score and the target vehicle's memory data is used to adjust the multi-dimensional weights of the target interaction scenario. It should be noted that if the result is correct, If the result is incorrect, the value is 1. The value is 0. Specifically, adjusting the multi-dimensional weights of the target interaction scenario can be done based on the loss function. The weights for each dimension are updated according to the following formula: ; Where d represents spatial dimension, temporal dimension, semantic dimension, or contextual dimension. Let η be the confidence level for the corresponding dimension and η be the learning rate. After the weights are updated, the weights for each dimension are normalized so that each weight is not less than 0 and the sum of the weights is 1.

[0093] Based on the target confidence level of the target vehicle memory data in the target interaction scenario, a three-level differentiated control strategy is implemented, rather than simply outputting the retrieval results directly: ; Direct Output (DIRECT_OUTPUT): With sufficient memory confidence, it is directly output as a deterministic result to downstream applications or user interfaces without any added uncertainty annotations; PROMPTED_OUTPUT: Since memory has a certain degree of uncertainty, the output result is accompanied by confidence level information, including the overall confidence level value and the dimension with the lowest confidence level (such as "there may be deviation in spatial location"), so that the user or downstream system can make its own judgment. Active confirmation (CONFIRM_REQUEST): If the memory is not reliable enough, the system will not output the result and will actively send a confirmation request to the user, who will then decide whether to accept it.

[0094] in, and It can be configured independently according to different application scenarios. Safety-sensitive scenarios (such as autonomous driving and valet parking) adopt higher threshold standards, while information display scenarios (such as driving log browsing) adopt more lenient threshold standards.

[0095] In this embodiment, a three-level differentiated access control strategy is used to divide the system output into three modes: direct output, prompting output, and active confirmation output. This enables the system to respond quickly when the data is fully reliable and to actively avoid risks when the data is questionable, effectively reducing practical application risks such as false alarms in voice interaction.

[0096] The technical solution of this embodiment involves acquiring input information; determining a target interaction scenario based on the input information; determining the target confidence level of the target vehicle memory data under the target interaction scenario based on the spatial confidence level, temporal confidence level, semantic confidence level, contextual confidence level of the target vehicle memory data corresponding to the input information, and the multi-dimensional weights of the target interaction scenario; inputting the target confidence level of the target vehicle memory data under the target interaction scenario, the input information, and the target vehicle memory data into a target model to obtain target control commands. This approach allows for the pre-determination of the target confidence level of the target vehicle memory data, and by inputting the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data into the target model to obtain target control commands, resulting in more accurate control commands.

[0097] Example 2 Figure 2 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0098] like Figure 2 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory 12 or a random access memory 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the read-only memory 12 or loaded from storage unit 18 into the random access memory 13. The random access memory 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, read-only memory 12, and random access memory 13 are interconnected via a bus 14. An input / output interface 15 is also connected to the bus 14.

[0099] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0100] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing units, graphics processing units, various special-purpose artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as data processing methods.

[0101] In some embodiments, the data processing method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via read-only memory 12 and / or communication unit 19. When the computer program is loaded into random access memory 13 and executed by processor 11, one or more steps of the data processing method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the data processing method by any other suitable means (e.g., by means of firmware).

[0102] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits (ASICs), application-specific standard products (ASICs), systems-on-a-chip (SoCs), payload programmable logic devices, computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0103] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0104] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0105] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a cathode ray tube or liquid crystal display) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0106] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0107] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to address the shortcomings of traditional physical hosts and virtual private servers, such as high management difficulty and weak business scalability.

[0108] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0109] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the data processing method according to any embodiment of the invention.

[0110] In the implementation of a computer program product, computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0111] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A data processing method, characterized in that, include: Obtain input information; Based on the input information, determine the target interaction scenario; Based on the spatial confidence, temporal confidence, semantic confidence, contextual confidence of the target vehicle memory data corresponding to the input information, and the multi-dimensional weights of the target interaction scenario, the target confidence of the target vehicle memory data under the target interaction scenario is determined. In the target interaction scenario, the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data are input into the target model to obtain the target control command.

2. The method according to claim 1, characterized in that, Based on the spatial confidence, temporal confidence, semantic confidence, contextual confidence, and multi-dimensional weights of the target vehicle memory data corresponding to the input information, and the target interaction scenario, the target confidence of the target vehicle memory data under the target interaction scenario is determined, including: Obtain the spatial confidence, temporal confidence, semantic confidence, and contextual confidence of the target vehicle memory data corresponding to the input information; The multi-dimensional weights of the target interaction scenario are obtained by querying a preset mapping table, wherein the multi-dimensional weights include: spatial weight, temporal weight, semantic weight, and contextual weight. The sum of the product of the spatial weight and the spatial confidence, the product of the temporal weight and the temporal confidence, the product of the semantic weight and the semantic confidence, and the product of the contextual weight and the contextual confidence is taken as the target confidence of the target vehicle memory data in the target interaction scenario.

3. The method according to claim 2, characterized in that, Obtaining the spatial confidence, temporal confidence, semantic confidence, and contextual confidence of the target vehicle memory data corresponding to the input information includes: Obtain the target vehicle memory data corresponding to the input information, wherein the target vehicle memory data includes: sensor data within a target time period, the target time period being prior to the time of receiving the input information; Based on the physical deviation of the target vehicle memory data, the initial weight of the positioning source corresponding to the target vehicle memory data, and the spatial coordinates of the positioning source, the spatial confidence of the target vehicle memory data is determined, wherein the positioning source corresponding to the target vehicle memory data is the location-related sensor data in the target vehicle memory data; Based on the target vehicle event type corresponding to the sensor data and the semantic association graph corresponding to the target vehicle event type, determine the semantic confidence of the target vehicle memory data; The contextual confidence of the target vehicle's memory data is determined based on the sensor data corresponding to the environmental perception channels in each environmental dimension of the target vehicle's memory data. The temporal confidence level of the target vehicle memory data is determined based on the target vehicle speed, target adaptive coefficient, and vehicle speed within the target window corresponding to the target vehicle memory data.

4. The method according to claim 3, characterized in that, Based on the physical deviation of the target vehicle memory data, the initial weight of the positioning source corresponding to the target vehicle memory data, and the spatial coordinates of the positioning source, the spatial confidence of the target vehicle memory data is determined, including: The target weight of the positioning source corresponding to the target vehicle memory data is determined based on the physical deviation of the target vehicle memory data and the initial weight of the positioning source corresponding to the target vehicle memory data. Based on the target weight of the location source corresponding to the target vehicle memory data and the spatial coordinates of the location source, the weighted consensus position of the location source is determined. The consensus deviation of the positioning source is determined based on the spatial coordinates of the positioning source and the weighted consensus position of the positioning source. The spatial consensus score of the target vehicle's memory data is determined based on the standard deviation of the consensus bias of the location source and the preset bias threshold. The spatial confidence level of the target vehicle's memory data is determined based on the spatial consensus score, physical deviation, and preset positioning accuracy parameters of the target vehicle's memory data.

5. The method according to claim 3, characterized in that, The target vehicle memory data includes: at least two sensor data corresponding to the target vehicle event type; Based on the target vehicle event type corresponding to the sensor data and the semantic association graph corresponding to the target vehicle event type, the semantic confidence of the target vehicle's memory data is determined, including: Determine the set of sensor data pairs based on at least two sensor data points corresponding to the target vehicle event type; Based on the semantic association graphs corresponding to the sensor data pairs in the sensor data pair set and the target vehicle event type, determine the number of union nodes, the number of intersection nodes, and the edge weights of the intersection nodes for each sensor data pair. The semantic score of the sensor data pair is determined based on the number of union nodes, the number of intersection nodes, and the edge weight of the target node. The semantic confidence of the target vehicle memory data is determined based on the semantic score of the sensor data pair and the amount of sensor data corresponding to the target vehicle event type.

6. The method according to claim 3, characterized in that, Based on the target vehicle speed, target adaptive coefficient, and vehicle speed within the target window corresponding to the target vehicle's memory data, determine the temporal confidence level of the target vehicle's memory data, including: The target window length is determined based on the target vehicle speed corresponding to the target vehicle's memory data. Based on the target window length, obtain the vehicle speed, the mean vehicle speed, and the standard deviation of the vehicle speed within the target window; The target adaptive coefficient is determined based on the spatial consensus score of the target vehicle's memory data, the initial adaptive coefficient, and the adjustment increment. If the absolute value of the difference between the vehicle speed in the target window and the mean vehicle speed in the target window is greater than the product of the target adaptive coefficient and the standard deviation of the vehicle speed in the target window, then the first value will be used as the time confidence of the target vehicle memory data. If the absolute value of the difference between the vehicle speed in the target window and the mean vehicle speed in the target window is less than or equal to the product of the target adaptive coefficient and the standard deviation of the vehicle speed in the target window, then the time confidence level of the target vehicle memory data is determined based on the target adaptive coefficient, the vehicle speed in the target window, and the adjustment parameters.

7. The method according to claim 3, characterized in that, Based on the sensor data corresponding to the environmental perception channels in each environmental dimension of the target vehicle's memory data, the contextual confidence of the target vehicle's memory data is determined, including: The contextual confidence level of the target vehicle's memory data is determined based on the total number of environmental dimensions of the target vehicle's memory data, the total number of environmental perception channels in each environmental dimension, the sensor data corresponding to each environmental perception channel, and the preset weights of each environmental dimension.

8. The method according to claim 1, characterized in that, In the target interaction scenario, the target confidence level of the target vehicle memory data, the input information, and the target vehicle memory data are input into the target model to obtain target control commands, including: Obtain the timestamp, spatial location vector, voice tag, and context tag of the target vehicle's memory data; The input information, the target vehicle memory data, the timestamp of the target vehicle memory data, the spatial location vector, the voice tag, the context tag, and the target confidence score of the target vehicle memory data under the target interaction scenario are input into the target model to obtain the target control command.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the data processing method according to any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the data processing method according to any one of claims 1-8.

11. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the data processing method according to any one of claims 1-8.

Images