A method and system for controlling in-vehicle atmosphere lighting based on voice analysis

Abstract

The present invention belongs to the technical field of automobile intelligent control. A method and system for controlling in-vehicle atmosphere lights based on voice analysis are provided. The method comprises: obtaining the first in-vehicle voice information of the vehicle through a pickup microphone, parsing the atmosphere light voice control instruction from the first in-vehicle voice information, and determining the atmosphere light target parameters corresponding to the atmosphere light voice control instruction; using a concentration intervention analysis model to evaluate the atmosphere light target parameters and obtain a concentration intervention index; obtaining the driving environment information of the vehicle, inputting the driving environment information into a driving risk prediction model, and obtaining a predicted driving risk value; using an atmosphere light control matching analysis model to comprehensively analyze the concentration intervention index and the driving risk value and obtain a matching index, if the matching index is higher than the matching threshold, the atmosphere light voice control instruction is executed, otherwise the atmosphere light voice control instruction is not executed. The scheme of the present invention can ensure driving safety during atmosphere light control.

Description

Technical Field

[0001] The present invention relates to the field of automobile intelligent control technology, and in particular to a method and system for controlling in-vehicle atmosphere lights based on voice analysis. Background Art

[0002] Car atmosphere lights are lighting devices installed inside cars to create a comfortable atmosphere. They usually use LEDs as light sources and achieve rich color changes through RGB color modes. In order to improve the convenience of adjusting car atmosphere lights and driving safety, many car manufacturers set the adjustment of car atmosphere lights to be realized by voice, that is, the user speaks the color, brightness, color switching frequency, etc. of the atmosphere lights that need to be adjusted, and then uses voice recognition to analyze the user's adjustment needs, thereby automatically realizing the adjustment of the color, brightness, color switching frequency, etc. of the atmosphere lights.

[0003] Although voice control can improve driving safety to a certain extent while ensuring the convenience of adjusting the ambient light, the unreasonable use of the ambient light can also bring about driving safety problems. For example, the user adjusts the brightness of the ambient light too brightly, or adjusts the brightness change frequency of the ambient light too quickly, resulting in an overly strong ambient light effect, which will affect the driver's driving concentration. The prior art lacks consideration of driving safety when using the ambient light, and the present invention is intended to solve or improve this technical problem.

[0004] It can be seen that the existing technology for voice adjustment of automobile ambient lights still has shortcomings in driving safety and needs to be improved. Summary of the invention

[0005] In this regard, the present invention provides a method, system, electronic device, computer storage medium and computer program product for controlling in-vehicle ambient light based on voice analysis to solve the above technical problems.

[0006] The present invention provides a method for controlling an in-vehicle ambient light based on voice analysis, the method comprising the following steps: obtaining first in-vehicle voice information of the vehicle through a pickup microphone, parsing the first in-vehicle voice information to obtain an ambient light voice control instruction, and determining an ambient light target parameter corresponding to the ambient light voice control instruction; evaluating the ambient light target parameter using a concentration intervention analysis model to obtain a concentration intervention index; obtaining driving environment information of the vehicle, the driving environment information including road structure information within a preset distance around the vehicle and real-time vehicle operating status information, inputting the driving environment information into a driving risk prediction model to obtain a predicted driving risk value; using an ambient light control matching analysis model to comprehensively analyze the concentration intervention index and the driving risk value to obtain a matching index, if the matching index is higher than a matching threshold, executing the ambient light voice control instruction, otherwise not executing the ambient light voice control instruction.

[0007] In some embodiments, the acquiring of the first in-car voice information of the vehicle through the sound pickup microphone includes: when the vehicle is started, the sound pickup microphone continuously acquires several groups of second in-car voice information of the vehicle until the driving speed of the vehicle is higher than the vehicle speed threshold; performing voiceprint analysis on each of the second in-car voice information, and performing voice classification; if the result of the voice classification is 1 person, adjusting the direction angle of the sound pickup microphone to align it with the driver's position; if the result of the voice classification is greater than 1 person, not adjusting the direction angle of the sound pickup microphone; acquiring the first in-car voice information of the vehicle through the sound pickup microphone.

[0008] In some embodiments, the use of a concentration intervention analysis model to evaluate the ambient light target parameters to obtain a concentration intervention index includes: inputting the ambient light target parameters into the concentration intervention analysis model, the concentration intervention analysis model outputting a preliminary concentration intervention index; obtaining a third in-car voice information of the vehicle through a pickup microphone, parsing the third in-car voice information to derive the intensity of in-car conversations involving the ambient light control content, wherein the in-car conversation intensity is comprehensively determined by the number of conversation participants, the frequency of conversations, and the height of the conversation pitch; wherein the third in-car voice information is earlier than the first in-car voice information; calculating an adjustment coefficient based on the in-car conversation intensity and the corresponding positive correlation, using the adjustment coefficient to adjust the preliminary concentration intervention index to obtain the final concentration intervention index.

[0009] In some embodiments, the obtaining of driving environment information of the vehicle, the driving environment information including road structure information within a preset distance around the vehicle and real-time vehicle operating status information, inputting the driving environment information into a driving risk prediction model to obtain a predicted driving risk value, includes: obtaining driving record data of the driver of the vehicle and accident record data of the vehicle, evaluating a bad driving index based on the driving record data and the accident record data, and determining a preset distance based on the bad driving index; obtaining road structure information within the preset distance around the vehicle and real-time vehicle operating status information, inputting the driving environment information into the driving risk prediction model to obtain a predicted driving risk value; wherein, the driving risk prediction model and the concentration intervention analysis model are both constructed based on Transformer.

[0010] In some embodiments, the use of the ambient light control matching analysis model to comprehensively analyze the concentration intervention index and the driving risk value to obtain a matching index includes: inputting the concentration intervention index and the driving risk value into the ambient light control matching analysis model, the ambient light control matching analysis model performing multiple rounds of simulated driving simulations based on the concentration intervention index and the driving risk value, and calculating an accident rate based on the results of each simulated driving simulation; and calculating the matching index based on the accident rate.

[0011] In some embodiments, the total number of rounds of the simulated driving simulation is determined by calculating the parameter difference between the ambient light target parameter and the ambient light current parameter, and determining the total number of rounds of the simulated driving simulation based on the parameter difference. Specifically, the parameter difference is positively correlated with the total number of rounds.

[0012] The present invention also provides an in-vehicle ambient light control system based on voice analysis, the system comprising a voice command parsing module, a concentration intervention analysis module, a driving risk analysis module, and a matching analysis module; the voice command parsing module is used to obtain the first in-vehicle voice information of the vehicle through a pickup microphone, parse the first in-vehicle voice information to obtain the ambient light voice control instruction, and determine the ambient light target parameter corresponding to the ambient light voice control instruction; the concentration intervention analysis module is used to evaluate the ambient light target parameter using a concentration intervention analysis model to obtain a concentration intervention index; the driving risk analysis module is used to obtain the driving environment information of the vehicle, the driving environment information includes road structure information within a preset distance around the vehicle and real-time vehicle operating status information, and input the driving environment information into a driving risk prediction model to obtain a predicted driving risk value; the matching analysis module is used to use the ambient light control matching analysis model to comprehensively analyze the concentration intervention index and the driving risk value to obtain a matching index, if the matching index is higher than a matching threshold, the ambient light voice control instruction is executed, otherwise the ambient light voice control instruction is not executed.

[0013] The present invention also provides an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement any of the methods described above.

[0014] The present invention also provides a computer storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement any of the above methods.

[0015] The present invention also provides a computer program product, characterized in that: the computer program includes a number of computer codes, and when the computer codes are executed by a processor, any of the above methods is implemented.

[0016] The present invention realizes a comprehensive analysis of whether the ambient light voice control instructions match the current driving environment through the mutual coordination of three pre-constructed models, namely, a concentration intervention analysis model, a driving risk prediction model and an ambient light control matching analysis model, so as to decide whether to execute the ambient light voice control instructions input by the occupants, thereby reducing driving safety problems caused by unreasonable use of ambient lights. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0018] Figure 1 It is a flow chart of a method for controlling in-vehicle atmosphere lighting based on voice analysis disclosed in an embodiment of the present invention.

[0019] Figure 2 It is a schematic diagram of the process of obtaining the concentration intervention index disclosed in an embodiment of the present invention.

[0020] Figure 3 It is a schematic diagram of the process of obtaining the driving risk value disclosed in the embodiment of the present invention.

[0021] Figure 4 It is a structural schematic diagram of a vehicle interior atmosphere light control system based on voice analysis disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following is a description of the implementation of the present application by specific specific embodiments. People familiar with the technology can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0023] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0024] like Figure 1As shown, an embodiment of the present invention discloses a method for controlling an in-vehicle ambient light based on voice analysis, the method comprising the following steps: obtaining first in-vehicle voice information of the vehicle through a pickup microphone, parsing the first in-vehicle voice information to obtain an ambient light voice control instruction, and determining an ambient light target parameter corresponding to the ambient light voice control instruction; evaluating the ambient light target parameter using a concentration intervention analysis model to obtain a concentration intervention index; obtaining driving environment information of the vehicle, the driving environment information including road structure information within a preset distance around the vehicle and real-time vehicle operating status information, inputting the driving environment information into a driving risk prediction model to obtain a predicted driving risk value; using an ambient light control matching analysis model to comprehensively analyze the concentration intervention index and the driving risk value to obtain a matching index, if the matching index is higher than a matching threshold, executing the ambient light voice control instruction, otherwise not executing the ambient light voice control instruction.

[0025] The present invention constructs the above three models, namely, the concentration intervention analysis model, the driving risk prediction model and the ambient light control matching analysis model. Based on the mutual cooperation of these three models, a comprehensive analysis is achieved on whether the ambient light voice control instructions match the current driving environment, so as to decide whether to execute the ambient light voice control instructions input by the occupants, thereby reducing driving safety problems caused by unreasonable use of ambient lights.

[0026] Specifically, the interior of the car is provided with pickup microphones, which are in the pickup state after the car is started. When the occupants need to adjust the ambient light, they speak voice information, and the pickup microphones obtain the voice information in the car. By performing voice analysis on the obtained in-car voice information, the corresponding ambient light voice control instructions can be obtained, and then the ambient light target parameters corresponding to the ambient light voice control instructions can be determined. The ambient light target parameters refer to the target brightness, target color, target color switching frequency and other control parameters of the ambient light. The pre-built concentration intervention analysis model is used to evaluate the ambient light target parameters, and it can be concluded that in the current environment, if the ambient light is adjusted according to the ambient light target parameters, the degree of intervention of the adjusted ambient light on the driver's driving concentration is expressed by the concentration intervention index. The higher the concentration intervention index, the greater the reduction in the driver's driving concentration.

[0027] Next, the driving environment information of the vehicle is obtained, mainly including the road structure information (number of lanes, bridge structure, lane line color, etc.) within a preset distance around the vehicle, and the real-time vehicle operation status information (vehicle speed, position, number of vehicles, etc.). The driving environment information is input into the pre-built driving risk prediction model to obtain the driving risk value predicted by the model. The driving risk value is used to represent the probability value of driving risk in the current driving environment. The higher the driving risk value, the higher the probability of driving risk.

[0028] Finally, the pre-built atmosphere light control matching analysis model is used to conduct a comprehensive analysis of the concentration intervention index and the driving risk value to obtain the corresponding matching index. The matching index is used to indicate the degree of matching between the atmosphere light voice control command output by the vehicle occupant and the current driving environment corresponding to the above-mentioned driving risk value. The larger the matching index, the higher the matching degree, and correspondingly, the higher the rationality of the atmosphere light voice control command. Therefore, when the matching index is higher than the matching threshold, it is determined that the switching adjustment of the atmosphere light will not affect driving safety, and the atmosphere light voice control command is executed; otherwise, it is determined that the switching adjustment of the atmosphere light will affect driving safety, and the atmosphere light voice control command is not executed.

[0029] In some embodiments, the acquiring of the first in-car voice information of the vehicle through the sound pickup microphone includes: when the vehicle is started, the sound pickup microphone continuously acquires several groups of second in-car voice information of the vehicle until the driving speed of the vehicle is higher than the vehicle speed threshold; performing voiceprint analysis on each of the second in-car voice information, and performing voice classification; if the result of the voice classification is 1 person, adjusting the direction angle of the sound pickup microphone to align it with the driver's position; if the result of the voice classification is greater than 1 person, not adjusting the direction angle of the sound pickup microphone; acquiring the first in-car voice information of the vehicle through the sound pickup microphone.

[0030] In this embodiment, the orientation angle of the pickup microphone is adjusted so that it is aimed at the occupant who needs to issue a voice command to adjust the atmosphere light, which can significantly improve the clarity of the first in-car voice information, thereby facilitating the identification of more accurate atmosphere light voice control instructions. In this regard, the present invention is arranged that when the vehicle is started, the pickup microphone continuously obtains several groups of second in-car voice information of the vehicle until the driving speed of the vehicle exceeds the vehicle speed threshold (for example, 5km / h), at which time it is impossible for new passengers to enter the vehicle, that is, the number of passengers in the vehicle is fixed. Then, the picked-up second in-car voice information is analyzed to obtain multiple voiceprint data, and these voiceprint data are classified by human voice, that is, it is determined how many passengers these voiceprint data come from. If the human voice classification result is 1 person, it means that there is only the driver in the vehicle, and there are no passengers in other seats. At this time, the pickup microphone can be directly controlled to change its orientation angle to align with the driver's position to ensure that the clearest in-car voice information can be obtained. If the result of the human voice classification is greater than one person, it means that there are multiple people in the car, and it is uncertain which passenger may output the in-car voice information for adjusting the ambient light. At this time, the direction angle of the pickup microphone is not adjusted.

[0031] In some embodiments, Figure 2 As shown, the use of the concentration intervention analysis model to evaluate the ambient light target parameters to obtain the concentration intervention index includes: inputting the ambient light target parameters into the concentration intervention analysis model, and the concentration intervention analysis model outputs a preliminary concentration intervention index; obtaining the third in-car voice information of the vehicle through a pickup microphone, and parsing the in-car conversation intensity involving the ambient light control content from the third in-car voice information, wherein the in-car conversation intensity is comprehensively determined by the number of conversation participants, conversation frequency, and conversation tone height; wherein the third in-car voice information is earlier than the first in-car voice information; and calculating an adjustment coefficient based on the in-car conversation intensity and the corresponding positive correlation, and using the adjustment coefficient to adjust the preliminary concentration intervention index to obtain the final concentration intervention index.

[0032] In this embodiment, the preset concentration intervention analysis model can perform a comprehensive analysis on the aforementioned obtained ambient light target parameters, thereby obtaining a preliminary concentration intervention index in the absence of other interference factors.

[0033] At the same time, the driver and other passengers in the car may have a high-intensity conversation about the control of the atmosphere light, such as the driver "enthusiastically" introducing the atmosphere light function of the car to other passengers, or other passengers "strongly" talking to the driver about the atmosphere light function of the car. When there is such a high-intensity conversation, it means that the driver will be more inclined to watch the atmosphere light after the control, so this factor will also affect the concentration intervention index. In this regard, before the passenger in the car outputs the first in-car voice information, the pickup microphone has acquired the third in-car voice information of the car, and the semantic understanding of these third in-car voice information can be analyzed to obtain the in-car conversation content involving the control of the atmosphere light, and the number of people in the conversation, the frequency of the conversation, and the height of the conversation tone can be further analyzed to calculate the corresponding conversation intensity. Obviously, the conversation intensity is positively correlated with the number of people in the conversation, the frequency of the conversation, and the height of the conversation tone. Then, the adjustment coefficient is calculated based on the in-car conversation intensity. The adjustment coefficient is, for example, a value greater than 1, and it is positively correlated with the conversation intensity, but the present invention does not limit the specific expression form of the positive correlation. Finally, the initial concentration intervention index is adjusted (e.g., multiplied) using the determined adjustment coefficient, and the concentration intervention index is finally obtained. Obviously, compared with the initial concentration intervention index, the final concentration intervention index takes into account the probability that the driver actively checks the adjusted ambient light, and its accuracy is higher.

[0034] In some embodiments, Figure 3 As shown, the driving environment information of the vehicle is obtained, and the driving environment information includes road structure information and real-time vehicle operation status information within a preset distance around the vehicle, and the driving environment information is input into the driving risk prediction model to obtain a predicted driving risk value, including: obtaining the driving record data of the driver of the vehicle and the accident record data of the vehicle, evaluating and deriving a bad driving index based on the driving record data and the accident record data, and determining a preset distance based on the bad driving index; obtaining the road structure information and real-time vehicle operation status information within the preset distance around the vehicle, and inputting the driving environment information into the driving risk prediction model to obtain the predicted driving risk value; wherein, the driving risk prediction model and the concentration intervention analysis model are both constructed based on Transformer.

[0035] In this embodiment, when the driving risk value of the vehicle is comprehensively analyzed based on the surrounding driving environment, the driving environment information of the vehicle cannot be obtained according to a fixed distance. The present invention sets a dynamic preset distance. Specifically, the driving record data of the driver of the vehicle and the accident record data of the vehicle are obtained, wherein the driving record data will include the current driver's bad driving behavior, such as sudden braking, sudden acceleration, love of changing lanes, frequent irregular use of turn signals, etc., and the accident record data contains information about traffic accidents that have occurred in the vehicle, and these traffic accident information may involve different drivers. It should be noted that after getting on the vehicle, the driver's identity can be determined by face, fingerprint and other recognition methods, and then the driving record data of the driver can be retrieved, and the accident record data of the vehicle is stored in the memory or server of the vehicle, and can also be retrieved.

[0036] Then, the bad driving index of the vehicle in the current situation is evaluated based on the driving record data and accident record data. The higher the bad driving index, the greater the probability of a traffic accident due to the vehicle. Correspondingly, the preset distance is set to be smaller, so as to focus on analyzing the driving risk between the vehicle and the vehicles at close range; and if the bad driving index is lower, the probability of a traffic accident due to the vehicle is lower. Correspondingly, the preset distance is set to be larger, so as to focus on analyzing the driving risk caused by the bad driving behavior of other vehicles around the vehicle. Moreover, the larger the preset distance, the more the overall driving risk value of the driving environment in which the vehicle is located can be analyzed.

[0037] Finally, the road structure information around the vehicle and the real-time vehicle operation status information are obtained according to the preset distance, and a pre-built driving risk prediction model is used for comprehensive analysis to obtain the predicted driving risk value.

[0038] It should be noted that the driving risk prediction model in the present invention and the above-mentioned concentration intervention analysis model are preferably constructed based on Transformer. The driving risk prediction model and the concentration intervention analysis model adopt the same model architecture, but need to be trained using corresponding training data. The constructed model includes main functional modules such as encoder, decoder, positional encoding, layer normalization, and residual connection. Among them, the encoder and decoder are composed of multiple identical layers, each of which includes two main sub-layers: Self-Attention Layer: allows the model to take into account the information of the entire sequence when processing each element, thereby capturing the dependencies within the sequence; Feed-Forward Neural Network: further nonlinear transformation of the output of the self-attention layer.

[0039] In some embodiments, the use of the ambient light control matching analysis model to comprehensively analyze the concentration intervention index and the driving risk value to obtain a matching index includes: inputting the concentration intervention index and the driving risk value into the ambient light control matching analysis model, the ambient light control matching analysis model performing multiple rounds of simulated driving simulations based on the concentration intervention index and the driving risk value, and calculating an accident rate based on the results of each simulated driving simulation; and calculating the matching index based on the accident rate.

[0040] In this embodiment, the atmosphere light control matching analysis model can perform multiple rounds of simulated driving simulations based on the aforementioned concentration intervention index and driving risk value. Specifically, a car simulation environment is pre-built through VISSIM software, and the car's driving parameters are configured based on the concentration intervention index, such as the probability of head deviation, speeding probability, etc., and the driving environment around the vehicle is configured based on the driving risk value, such as the number of vehicles, vehicle distribution position, vehicle speed, etc. After each simulated driving simulation, the simulated driving simulation result of whether the vehicle has a traffic accident is obtained, and then the accident rate is obtained by dividing the accident result by the total simulation rounds. Finally, the above-mentioned matching index is calculated based on the accident rate, for example, matching index = 1-accident rate * conversion coefficient.

[0041] Among them, the atmosphere light control matching analysis model includes an input layer, a simulation layer, a matching analysis layer and an output layer. The input layer is used to receive the aforementioned concentration intervention index and driving risk value, the simulation layer is used to perform multiple rounds of simulation based on the above-mentioned automobile simulation environment, the matching analysis layer is used to calculate the accident rate and matching index based on the simulated driving simulation results, and the output layer outputs the matching index. The atmosphere light control matching analysis model can be built based on a convolutional neural network or the above-mentioned Transformer, and the present invention does not make specific limitations.

[0042] It should be noted that the car simulation environment is pre-built and does not need to be built in real time, which can greatly reduce the time consumption of matching analysis and thus improve response efficiency. Moreover, the car simulation environment is lightweight and other unnecessary content is deleted, which can further improve response efficiency. In addition, the above simulation can be carried out in multiple rounds simultaneously, which can further shorten the simulation time.

[0043] In some embodiments, the total number of rounds of the simulated driving simulation is determined by calculating the parameter difference between the ambient light target parameter and the ambient light current parameter, and determining the total number of rounds of the simulated driving simulation based on the parameter difference. Specifically, the parameter difference is positively correlated with the total number of rounds.

[0044] In this embodiment, the parameter difference between the target parameter of the ambient light and the current parameter of the ambient light is first calculated, and the parameter difference is, for example, the brightness difference, the color switching frequency difference, etc. In addition, the prominence of different light colors can be pre-determined, and then the above parameter difference between the different template ambient light colors and the current ambient light color is determined based on the prominence difference (mainly for the enhancement of the parameter difference, such as increased brightness, more significant color, higher color switching frequency, etc.). The total rounds of simulated driving simulation are determined according to the parameter difference and the corresponding positive correlation. When the parameter difference is larger, the concentration intervention index is generally larger, and the probability of mismatch between the concentration intervention index and the driving risk value is also greater. At this time, more rounds of simulated driving simulation are set to obtain a more accurate accident rate; on the contrary, fewer rounds of simulated driving simulation are set, so that the response speed when the ambient light parameters are slightly adjusted can be achieved, that is, the analysis speed of the matching index after the simulated driving simulation is reduced becomes faster, so that the ambient light voice control command can be responded to and executed faster to reduce the frustration of the ambient light control.

[0045] like Figure 4As shown, an embodiment of the present invention further discloses a vehicle interior ambient light control system based on voice analysis, the system comprising a voice command parsing module, a concentration intervention analysis module, a driving risk analysis module, and a matching analysis module; the voice command parsing module is used to obtain a first in-vehicle voice information of the vehicle through a pickup microphone, parse the first in-vehicle voice information to obtain an ambient light voice control instruction, and determine the ambient light target parameter corresponding to the ambient light voice control instruction; the concentration intervention analysis module is used to evaluate the ambient light target parameter using a concentration intervention analysis model to obtain a concentration intervention index; the driving risk analysis module is used to obtain the driving environment information of the vehicle, the driving environment information includes road structure information within a preset distance around the vehicle and real-time vehicle operation status information, and input the driving environment information into a driving risk prediction model to obtain a predicted driving risk value; the matching analysis module is used to use the ambient light control matching analysis model to comprehensively analyze the concentration intervention index and the driving risk value to obtain a matching index, and if the matching index is higher than a matching threshold, the ambient light voice control instruction is executed, otherwise the ambient light voice control instruction is not executed.

[0046] An embodiment of the present invention further discloses an electronic device, comprising: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method described in the above embodiment.

[0047] An embodiment of the present invention further discloses a computer storage medium, wherein the computer storage medium stores a computer program, and the computer program is executed by a processor to implement the method described in the above embodiment.

[0048] The embodiment of the present invention further discloses a computer program product, characterized in that: the computer program includes a number of computer codes, and when the computer codes are executed by a processor, any of the above methods is implemented.

[0049] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0050] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0051] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for controlling in-vehicle atmosphere lights based on speech analysis, characterized in that: The method comprises the following steps: obtaining first in-vehicle voice information of the vehicle through a pickup microphone, parsing the first in-vehicle voice information to obtain an ambient light voice control instruction, and determining an ambient light target parameter corresponding to the ambient light voice control instruction; The ambient light target parameters are evaluated using a concentration intervention analysis model to obtain a concentration intervention index; the driving environment information of the vehicle is obtained, the driving environment information including road structure information within a preset distance around the vehicle and real-time vehicle operation status information, and the driving environment information is input into a driving risk prediction model to obtain a predicted driving risk value; The concentration intervention index and the driving risk value are comprehensively analyzed using the ambient light control matching analysis model to obtain a matching index. If the matching index is higher than a matching threshold, the ambient light voice control instruction is executed; otherwise, the ambient light voice control instruction is not executed.

2. The method for controlling the in-vehicle atmosphere light based on voice analysis according to claim 1, characterized in that: The first in-car voice information of the vehicle is obtained through the sound pickup microphone, including: when the vehicle is started, the sound pickup microphone continuously obtains several groups of second in-car voice information of the vehicle until the driving speed of the vehicle is higher than the vehicle speed threshold; each of the second in-car voice information is subjected to voiceprint analysis, and human voice classification is performed; if the human voice classification result is 1 person, the direction angle of the sound pickup microphone is adjusted to align it with the driver's position; if the human voice classification result is greater than 1 person, the direction angle of the sound pickup microphone is not adjusted; the first in-car voice information of the vehicle is obtained through the sound pickup microphone.

3. The method for controlling the in-vehicle atmosphere light based on voice analysis according to claim 1, characterized in that: The atmosphere light target parameters are evaluated using a concentration intervention analysis model to obtain a concentration intervention index, including: inputting the atmosphere light target parameters into the concentration intervention analysis model, and the concentration intervention analysis model outputs a preliminary concentration intervention index; obtaining a third in-car voice information of the vehicle through a pickup microphone, and parsing the third in-car voice information to obtain an in-car conversation intensity involving the atmosphere light control content, wherein the in-car conversation intensity is comprehensively determined by the number of conversation participants, the conversation frequency, and the conversation pitch height; wherein the third in-car voice information is earlier than the first in-car voice information; and calculating an adjustment coefficient based on the in-car conversation intensity and a corresponding positive correlation, and using the adjustment coefficient to adjust the preliminary concentration intervention index to obtain a final concentration intervention index.

4. The method for controlling the in-vehicle ambient light based on voice analysis according to claim 3, characterized in that: Acquire driving environment information of the vehicle, the driving environment information including road structure information and real-time vehicle operation status information within a preset distance around the vehicle, input the driving environment information into a driving risk prediction model, and obtain a predicted driving risk value, including: acquiring driving record data of the driver of the vehicle and accident record data of the vehicle, evaluating and deriving a bad driving index based on the driving record data and the accident record data, and determining a preset distance based on the bad driving index; acquire road structure information and real-time vehicle operation status information within the preset distance around the vehicle, input the driving environment information into the driving risk prediction model, and obtain a predicted driving risk value; wherein, the driving risk prediction model and the concentration intervention analysis model are both constructed based on Transformer.

5. The method for controlling the in-vehicle atmosphere light based on voice analysis according to claim 4, characterized in that: The concentration intervention index and the driving risk value are comprehensively analyzed using an ambient light control matching analysis model to obtain a matching index, including: inputting the concentration intervention index and the driving risk value into the ambient light control matching analysis model, the ambient light control matching analysis model performing multiple rounds of simulated driving simulations based on the concentration intervention index and the driving risk value, and calculating an accident rate based on the results of each simulated driving simulation; and calculating the matching index based on the accident rate.

6. The method for controlling the in-vehicle ambient light based on voice analysis according to claim 5, characterized in that: The total number of rounds of the simulated driving simulation is determined by calculating the parameter difference between the ambient light target parameter and the ambient light current parameter, and determining the total number of rounds of the simulated driving simulation based on the parameter difference. Specifically, the parameter difference is positively correlated with the total number of rounds.

7. A vehicle interior atmosphere light control system based on voice analysis, the system comprising a voice command parsing module, a concentration intervention analysis module, a driving risk analysis module, and a matching analysis module; characterized in that: The voice command parsing module is used to obtain first in-vehicle voice information of the vehicle through a pickup microphone, derive an ambient light voice control instruction from the first in-vehicle voice information, and determine an ambient light target parameter corresponding to the ambient light voice control instruction; The concentration intervention analysis module is used to evaluate the ambient light target parameters using the concentration intervention analysis model to obtain a concentration intervention index; the driving risk analysis module is used to obtain driving environment information of the vehicle, the driving environment information includes road structure information within a preset distance around the vehicle and real-time vehicle operation status information, and input the driving environment information into the driving risk prediction model to obtain a predicted driving risk value; The matching analysis module is used to use the ambient light control matching analysis model to comprehensively analyze the concentration intervention index and the driving risk value to obtain a matching index. If the matching index is higher than a matching threshold, the ambient light voice control instruction is executed; otherwise, the ambient light voice control instruction is not executed.

8. An electronic device, comprising: At least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the method according to any one of claims 1 to 6.

9. A computer storage medium storing a computer program, characterized in that: The computer program is executed by a processor to implement the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that: The computer program includes a number of computer codes, and when the computer codes are executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images